Agent hierarchy over the Jube service layer

Generated from the catalogue, not hand-listed: every agent below owns tools that exist in ServiceToolCatalogue, every change appears exactly once in the whole tree, and no agent holds more than fifteen. A read may appear more than once, because an agent that can change an object holds the reads of that object as well – the alternative was an author whose only way to look anything up was free-form SQL against a schema it had to guess.

Every operation is also listed flat, one entry each, in tool-inventory.md. That inventory is the backstop for this tree: when routing by meaning fails, an exact operation name, a service name or a permission number resolves there instead.

Shape


tools assigned	1046, over 718 distinct operations
leaf agents	166
supervisors	33
leaves attached straight to a domain	10
domains	9
delegation hops	root → domain → supervisor → leaf is three, exactly `RunContext.MaxDepth`; a collapsed slice is two
widest fan-out	9, counting cross-domain reach; `RunContext.MaxFanOut` caps one spawn call at 8, not the roster
largest toolbox	15
shared reads	328 assignments, all idempotent and non-destructive; no change is shared

Cross-domain reach

A leaf owns an operation outright and cannot delegate, so an agent whose job needs an operation filed in another slice has no route to it unless one is named. Where that happened the failure was never a refusal: the agent routed correctly, worked out exactly what it needed, found it was not holding it, and answered from whatever it could reach instead – a rule author quoting a documentation example as if it were the caller’s own schema, a datasource author hanging its work on ExampleVisualisation because it could not create the registry it meant.

Reach into another slice is widened by an edge, never by giving a second agent the same tool, because every change is owned by exactly one leaf and that is what makes the inventory below an address book. Sharing is confined to reads of an agent’s own slice, which is not reach at all: it is the author of an object being able to see the object it is about to change. The edges are declared in one table in AgentDefinitions.cs (CrossDomainReach), each carrying the sentence the agent is told, and they are marked Cross-Domain Reach on each agent below. Four kinds recur:

the completions leaf, which is the only authority on which payload fields a model actually carries and on which operators the rule builder offers – reached by every domain that authors something written against a field name;
the invocation context leaf, which builds the rule-evaluation context that every test bench’s Execute operation takes and none of them can build;
the synchronisation schedule leaf, which is the closing step of all nine rule taxonomy graphs and where every rule fragment is compiled, filed under platform reliability rather than beside the rules;
sibling slices of one artefact, where a datasource needs the registry it hangs off and the parameters it binds to.

Domains follow the application’s own navigation menu in _Layout.cshtml, because that is hand-authored information architecture with the permission gate already attached to each entry. A slice is split three ways into Inspector, Test Bench and Author only where it exceeds seven tools; smaller slices stay whole, and small related slices are packed together up to the same budget. The split is by what an agent may change, not by what it may see: the Author of a slice holds that slice’s whole read surface too, including its validation, and the Inspector remains the safe default because it holds no change at all.

Root

Jooby

Hierarchy Placement: Level 1, root
Semantic Persona: The entry point for every request. It owns no tools of its own and exists to route a question to the one domain that owns the data behind it.
Semantic Routing Keywords: (routes everything)
Toolbox (Max 15): none — routing only
Downstream Dependencies: access-admin, assistant-admin, case-analyst, case-workflow-admin, data-analyst, model-admin, model-risk, model-support, platform-ops

Access Control and Tenancy (`access-admin`)

Access Control and Tenancy

Hierarchy Placement: Level 2, child of Jooby
Semantic Persona: Owns identity, roles, permissions, API keys and tenancy. Deliberately separated from every functional domain so that the agent which grants access is never the agent that uses it.
Semantic Routing Keywords: user, role, permission, tenant, API key, login, logout, entitlement, segregation of duties, access
Toolbox (Max 15): none — routes to 3 supervisors covering 45 tools
Downstream Dependencies: access-admin.entitlement, access-admin.identity, access-admin.tenancy

Entitlement Supervisor

Hierarchy Placement: Level 3, child of Access Control and Tenancy
Semantic Persona: Routes within the entitlement slice of access control and tenancy. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: entitlement, role, registry, permission, specification
Toolbox (Max 15): none — routes to 7 leaf agents covering 19 tools
Downstream Dependencies: role-registry-inspector, role-registry-test-bench, role-registry-author, role-registry-permission-inspector, role-registry-permission-test-bench, role-registry-permission-author, permission-specification

Role Registry Inspector (`role-registry-inspector`)

Hierarchy Placement: Level 4, child of Entitlement Supervisor
Semantic Persona: Owns the role registry slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: role, registry, inspector
Toolbox (Max 15):
- RoleRegistryList: Lists Roles visible to the caller’s tenant, keyset-paged and capped. (permissions from RoleRegistryService.ListPermissions)
- RoleRegistryFilterFields: Lists the fields a query builder JSON filter over Role Registrys may use. ( permissions from RoleRegistryService.ListPermissions)
- RoleRegistryFilter: Returns the Role Registrys matching query builder JSON, keyset-paged and capped. ( permissions from RoleRegistryService.ListPermissions)
- RoleRegistryCount: Counts the Role Registrys matching query builder JSON, optionally grouped by a field. ( permissions from RoleRegistryService.ListPermissions)
- RoleRegistryGet: Returns a single Role by Id. (permissions from RoleRegistryService.RequiredPermissions)
- RoleRegistryGetByGuid: Returns a single Role by Id. (permissions from RoleRegistryService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Role Registry Test Bench (`role-registry-test-bench`)

Hierarchy Placement: Level 4, child of Entitlement Supervisor
Semantic Persona: Owns the role registry slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: role, registry, test, bench
Toolbox (Max 15):
- RoleRegistryValidate: Validates a Role Registry without saving it and returns every failure. (permissions from RoleRegistryService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Role Registry Author (`role-registry-author`)

Hierarchy Placement: Level 4, child of Entitlement Supervisor
Semantic Persona: Owns the role registry slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: role, registry, author
Toolbox (Max 15):
- RoleRegistryList: Lists Roles visible to the caller’s tenant, keyset-paged and capped. (permissions from RoleRegistryService.ListPermissions)
- RoleRegistryFilterFields: Lists the fields a query builder JSON filter over Role Registrys may use. ( permissions from RoleRegistryService.ListPermissions)
- RoleRegistryFilter: Returns the Role Registrys matching query builder JSON, keyset-paged and capped. ( permissions from RoleRegistryService.ListPermissions)
- RoleRegistryCount: Counts the Role Registrys matching query builder JSON, optionally grouped by a field. ( permissions from RoleRegistryService.ListPermissions)
- RoleRegistryGet: Returns a single Role by Id. (permissions from RoleRegistryService.RequiredPermissions)
- RoleRegistryGetByGuid: Returns a single Role by Id. (permissions from RoleRegistryService.RequiredPermissions)
- RoleRegistryValidate: Validates a Role Registry without saving it and returns every failure. (permissions from RoleRegistryService.RequiredPermissions)
- RoleRegistryCreate: Creates a Role; calling twice creates two. Write — confirmation required (permissions from RoleRegistryService.RequiredPermissions)
- RoleRegistryUpdate: Updates a Role; overwrites, writes a version-audit row. Write — confirmation required ( permissions from RoleRegistryService.RequiredPermissions)
- RoleRegistryDelete: Soft-deletes a Role; reversible only by direct database action. Write — confirmation required (permissions from RoleRegistryService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Role Registry Permission Inspector (`role-registry-permission-inspector`)

Hierarchy Placement: Level 4, child of Entitlement Supervisor
Semantic Persona: Owns the role registry permission slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: role, registry, permission, inspector
Toolbox (Max 15):
- RoleRegistryPermissionList: Lists Role Registry Permission grants visible to the caller’s tenant, keyset-paged and capped. (permissions from RoleRegistryPermissionService.RequiredPermissions)
- RoleRegistryPermissionFilterFields: Lists the fields a query builder JSON filter over Role Registry Permissions may use. (permissions from RoleRegistryPermissionService.RequiredPermissions)
- RoleRegistryPermissionFilter: Returns the Role Registry Permissions matching query builder JSON, keyset-paged and capped. (permissions from RoleRegistryPermissionService.RequiredPermissions)
- RoleRegistryPermissionCount: Counts the Role Registry Permissions matching query builder JSON, optionally grouped by a field. (permissions from RoleRegistryPermissionService.RequiredPermissions)
- RoleRegistryPermissionGet: Returns a single Role Registry Permission grant by Id. (permissions from RoleRegistryPermissionService.RequiredPermissions)
- RoleRegistryPermissionGetByGuid: Returns a single Role Registry Permission grant by Id. (permissions from RoleRegistryPermissionService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Role Registry Permission Test Bench (`role-registry-permission-test-bench`)

Hierarchy Placement: Level 4, child of Entitlement Supervisor
Semantic Persona: Owns the role registry permission slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: role, registry, permission, test, bench
Toolbox (Max 15):
- RoleRegistryPermissionValidate: Validates a Role Registry Permission without saving it and returns every failure. (permissions from RoleRegistryPermissionService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Role Registry Permission Author (`role-registry-permission-author`)

Hierarchy Placement: Level 4, child of Entitlement Supervisor
Semantic Persona: Owns the role registry permission slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: role, registry, permission, author
Toolbox (Max 15):
- RoleRegistryPermissionList: Lists Role Registry Permission grants visible to the caller’s tenant, keyset-paged and capped. (permissions from RoleRegistryPermissionService.RequiredPermissions)
- RoleRegistryPermissionFilterFields: Lists the fields a query builder JSON filter over Role Registry Permissions may use. (permissions from RoleRegistryPermissionService.RequiredPermissions)
- RoleRegistryPermissionFilter: Returns the Role Registry Permissions matching query builder JSON, keyset-paged and capped. (permissions from RoleRegistryPermissionService.RequiredPermissions)
- RoleRegistryPermissionCount: Counts the Role Registry Permissions matching query builder JSON, optionally grouped by a field. (permissions from RoleRegistryPermissionService.RequiredPermissions)
- RoleRegistryPermissionGet: Returns a single Role Registry Permission grant by Id. (permissions from RoleRegistryPermissionService.RequiredPermissions)
- RoleRegistryPermissionGetByGuid: Returns a single Role Registry Permission grant by Id. (permissions from RoleRegistryPermissionService.RequiredPermissions)
- RoleRegistryPermissionValidate: Validates a Role Registry Permission without saving it and returns every failure. (permissions from RoleRegistryPermissionService.RequiredPermissions)
- RoleRegistryPermissionCreate: Grants a Permission Specification to a Role Registry; calling twice creates two grants. Write — confirmation required (permissions from RoleRegistryPermissionService.RequiredPermissions)
- RoleRegistryPermissionUpdate: Updates a Role Registry Permission grant; writes a version-audit row. Write — confirmation required (permissions from RoleRegistryPermissionService.RequiredPermissions)
- RoleRegistryPermissionDelete: Soft-deletes a Role Registry Permission grant; reversible only by direct database action. Write — confirmation required (permissions from RoleRegistryPermissionService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Permission Specification (`permission-specification`)

Hierarchy Placement: Level 4, child of Entitlement Supervisor
Semantic Persona: Owns the permission specification slice end to end, which is small enough that reads and writes stay with one agent.
Semantic Routing Keywords: permission, specification
Toolbox (Max 15):
- PermissionSpecificationList: Lists rows in the global Permission Specification catalogue, keyset-paged and capped. (permissions from PermissionSpecificationService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Identity Supervisor

Hierarchy Placement: Level 3, child of Access Control and Tenancy
Semantic Persona: Routes within the identity slice of access control and tenancy. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: identity, user, registry, in, tenant, api, key, login, logout
Toolbox (Max 15): none — routes to 4 leaf agents covering 17 tools
Downstream Dependencies: user-registry-inspector, user-registry-test-bench, user-registry-author, user-observability

User Registry Inspector (`user-registry-inspector`)

Hierarchy Placement: Level 4, child of Identity Supervisor
Semantic Persona: Owns the user registry slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: user, registry, inspector
Toolbox (Max 15):
- UserRegistryList: Lists user accounts for the caller’s tenant, keyset-paged and capped. (permissions from UserRegistryService.RequiredPermissions)
- UserRegistryFilterFields: Lists the fields a query builder JSON filter over User Registrys may use. ( permissions from UserRegistryService.RequiredPermissions)
- UserRegistryFilter: Returns the User Registrys matching query builder JSON, keyset-paged and capped. ( permissions from UserRegistryService.RequiredPermissions)
- UserRegistryCount: Counts the User Registrys matching query builder JSON, optionally grouped by a field. ( permissions from UserRegistryService.RequiredPermissions)
- UserRegistryGet: Returns one user account by id, scoped to the caller’s tenant. (permissions from UserRegistryService.RequiredPermissions)
- UserRegistryGetByGuid: Returns one user account by Guid, scoped to the caller’s tenant. (permissions from UserRegistryService.RequiredPermissions)
- UserRegistryListByRoleRegistryGuid: Lists the user accounts assigned to a given Role Registry, scoped to the caller’s tenant. (permissions from UserRegistryService.RequiredPermissions)
Downstream Dependencies: none (leaf)

User Registry Test Bench (`user-registry-test-bench`)

Hierarchy Placement: Level 4, child of Identity Supervisor
Semantic Persona: Owns the user registry slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: user, registry, test, bench
Toolbox (Max 15):
- UserRegistryValidate: Validates an User Registry without saving it and returns every failure. (permissions from UserRegistryService.RequiredPermissions)
Downstream Dependencies: none (leaf)

User Registry Author (`user-registry-author`)

Hierarchy Placement: Level 4, child of Identity Supervisor
Semantic Persona: Owns the user registry slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: user, registry, author
Toolbox (Max 15):
- UserRegistryList: Lists user accounts for the caller’s tenant, keyset-paged and capped. (permissions from UserRegistryService.RequiredPermissions)
- UserRegistryFilterFields: Lists the fields a query builder JSON filter over User Registrys may use. ( permissions from UserRegistryService.RequiredPermissions)
- UserRegistryFilter: Returns the User Registrys matching query builder JSON, keyset-paged and capped. ( permissions from UserRegistryService.RequiredPermissions)
- UserRegistryCount: Counts the User Registrys matching query builder JSON, optionally grouped by a field. ( permissions from UserRegistryService.RequiredPermissions)
- UserRegistryGet: Returns one user account by id, scoped to the caller’s tenant. (permissions from UserRegistryService.RequiredPermissions)
- UserRegistryGetByGuid: Returns one user account by Guid, scoped to the caller’s tenant. (permissions from UserRegistryService.RequiredPermissions)
- UserRegistryListByRoleRegistryGuid: Lists the user accounts assigned to a given Role Registry, scoped to the caller’s tenant. (permissions from UserRegistryService.RequiredPermissions)
- UserRegistryValidate: Validates an User Registry without saving it and returns every failure. (permissions from UserRegistryService.RequiredPermissions)
- UserRegistryCreate: Registers a user account under a Role Registry in the caller’s tenant; calling twice creates two. Write — confirmation required (permissions from UserRegistryService.RequiredPermissions)
- UserRegistryUpdate: Updates a user account in the caller’s tenant by id. Never touches the password. Write — confirmation required (permissions from UserRegistryService.RequiredPermissions)
- UserRegistryDelete: Soft-deletes a user account in the caller’s tenant by id. Write — confirmation required (permissions from UserRegistryService.RequiredPermissions)
Downstream Dependencies: none (leaf)

User Observability (`user-observability`)

Hierarchy Placement: Level 4, child of Identity Supervisor
Semantic Persona: Owns the user observability slice end to end, which is small enough that reads and writes stay with one agent. It holds 2 mutating tools, so a confirmation is required before any of them runs.
Semantic Routing Keywords: user, in, tenant, registry, api, key, login, logout
Toolbox (Max 15):
- UserInTenantList: Lists the users allocated to the caller’s own tenant, keyset-paged and capped. (permissions from UserInTenantService.RequiredPermissions)
- UserInTenantGetCurrentTenantRegistry: Returns the caller’s own tenant allocation, or nothing if unallocated. ( permissions from UserInTenantService.RequiredPermissions)
- UserInTenantSwitchTenant: Switches the caller to a different tenant; restricted to a Landlord caller. Write — confirmation required (permissions from UserInTenantService.RequiredPermissions)
- UserRegistryApiKeyListByUserRegistryId: Lists the API keys issued against a UserRegistry, masked to an 8-character display prefix. (permissions from UserRegistryApiKeyService.RequiredPermissions)
- UserRegistryApiKeyDelete: Revokes an API key; reversible only by direct database action. A repeat call on an already-revoked key succeeds as a no-op if it was fully removed, or is reported not-found if it belongs to another tenant. Write — confirmation required (permissions from UserRegistryApiKeyService.RequiredPermissions)
- UserLoginList: Lists the Login Audit Trail – every sign-in attempt, successful or not, across all three schemes (Username/Password, Negotiate, OAuth/OpenID Connect), most recent first, capped at 100000, with optional date-range and substring search (Crea… (permissions from UserLoginService.RequiredPermissions)
- UserLogoutList: Lists the Logout Audit Trail – every time a session was cut (the user logged out, an administrator revoked the user’s tokens, or a password change revoked the earlier sessions), most recent first, capped at 100000, with optional date-range… (permissions from UserLogoutService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Tenancy Supervisor

Hierarchy Placement: Level 3, child of Access Control and Tenancy
Semantic Persona: Routes within the tenancy slice of access control and tenancy. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: tenancy, tenant, registry
Toolbox (Max 15): none — routes to 3 leaf agents covering 9 tools
Downstream Dependencies: tenant-registry-inspector, tenant-registry-test-bench, tenant-registry-author

Tenant Registry Inspector (`tenant-registry-inspector`)

Hierarchy Placement: Level 4, child of Tenancy Supervisor
Semantic Persona: Owns the tenant registry slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: tenant, registry, inspector
Toolbox (Max 15):
- TenantRegistryList: Lists registered Tenants, Id-ascending, keyset-paged and capped. Landlord-only. ( permissions from TenantRegistryService.RequiredPermissions)
- TenantRegistryFilterFields: Lists the fields a query builder JSON filter over Tenant Registrys may use. ( permissions from TenantRegistryService.RequiredPermissions)
- TenantRegistryFilter: Returns the Tenant Registrys matching query builder JSON, keyset-paged and capped. ( permissions from TenantRegistryService.RequiredPermissions)
- TenantRegistryCount: Counts the Tenant Registrys matching query builder JSON, optionally grouped by a field. ( permissions from TenantRegistryService.RequiredPermissions)
- TenantRegistryGet: Returns a single Tenant by Id. Landlord-only. (permissions from TenantRegistryService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Tenant Registry Test Bench (`tenant-registry-test-bench`)

Hierarchy Placement: Level 4, child of Tenancy Supervisor
Semantic Persona: Owns the tenant registry slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: tenant, registry, test, bench
Toolbox (Max 15):
- TenantRegistryValidate: Validates a Tenant Registry without saving it and returns every failure. (permissions from TenantRegistryService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Tenant Registry Author (`tenant-registry-author`)

Hierarchy Placement: Level 4, child of Tenancy Supervisor
Semantic Persona: Owns the tenant registry slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: tenant, registry, author
Toolbox (Max 15):
- TenantRegistryList: Lists registered Tenants, Id-ascending, keyset-paged and capped. Landlord-only. ( permissions from TenantRegistryService.RequiredPermissions)
- TenantRegistryFilterFields: Lists the fields a query builder JSON filter over Tenant Registrys may use. ( permissions from TenantRegistryService.RequiredPermissions)
- TenantRegistryFilter: Returns the Tenant Registrys matching query builder JSON, keyset-paged and capped. ( permissions from TenantRegistryService.RequiredPermissions)
- TenantRegistryCount: Counts the Tenant Registrys matching query builder JSON, optionally grouped by a field. ( permissions from TenantRegistryService.RequiredPermissions)
- TenantRegistryGet: Returns a single Tenant by Id. Landlord-only. (permissions from TenantRegistryService.RequiredPermissions)
- TenantRegistryValidate: Validates a Tenant Registry without saving it and returns every failure. (permissions from TenantRegistryService.RequiredPermissions)
- TenantRegistryCreate: Creates a Tenant; calling twice creates two. Landlord-only. Write — confirmation required (permissions from TenantRegistryService.RequiredPermissions)
- TenantRegistryUpdate: Updates a Tenant; overwrites, writes a version-audit row. Landlord-only. Write — confirmation required (permissions from TenantRegistryService.RequiredPermissions)
- TenantRegistryDelete: Soft-deletes a Tenant; reversible only by direct database action. Landlord-only. Write — confirmation required (permissions from TenantRegistryService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Assistant Conversation Administration (`assistant-admin`)

Assistant Conversation Administration

Hierarchy Placement: Level 2, child of Jooby
Semantic Persona: Administers the assistant’s own stored conversations and retrieval corpus. Strictly housekeeping: it can list and search what the assistant has recorded, never start a turn or spawn another agent.
Semantic Routing Keywords: conversation history, transcript, retrieval corpus, RAG chunk, assistant housekeeping
Toolbox (Max 15): none — routes to 1 leaf agent covering 5 tools
Downstream Dependencies: ask-jooby-observability

Ask Jooby Observability (`ask-jooby-observability`)

Hierarchy Placement: Level 3, child of Assistant Conversation Administration
Semantic Persona: Owns the ask jooby observability slice end to end, which is small enough that reads and writes stay with one agent. It holds 2 mutating tools, so a confirmation is required before any of them runs.
Semantic Routing Keywords: ask, jooby, conversation, rag, chunk
Toolbox (Max 15):
- AskJoobyConversationList: Lists AskJooby conversations for the caller’s tenant, most recently active first. ( permissions from AskJoobyConversationService.RequiredPermissions)
- AskJoobyConversationCreate: Starts a new, empty, untitled AskJooby conversation in the caller’s tenant. Write — confirmation required (permissions from AskJoobyConversationService.RequiredPermissions)
- AskJoobyConversationGetMessages: Returns the full transcript of one conversation, oldest first. (permissions from AskJoobyConversationService.RequiredPermissions)
- AskJoobyConversationSendMessage: Sends a message to Jooby and streams the reply, persisting both sides of the exchange. Write — confirmation required (permissions from AskJoobyConversationService.RequiredPermissions)
- AskJoobyRagChunkSearch: Searches the Pasture and returns the most relevant Bites, ranked by reciprocal-rank fusion of full-text and vector search. (permissions from AskJoobyRagChunkService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Case Investigation Desk (`case-analyst`)

Case Investigation Desk

Hierarchy Placement: Level 2, child of Jooby
Semantic Persona: Works live alert and case material for an investigator: the case itself, its notes, files, events and transaction journal, the activation watcher and sanctions screening. Never changes how the platform detects anything, only what is recorded about a case under investigation.
Semantic Routing Keywords: case, alert, investigate, SAR, escalate, note, attachment, journal, watcher, sanctions screening, disposition
Toolbox (Max 15): none — routes to 1 supervisor and 3 leaf agents covering 29 tools
Downstream Dependencies: case-analyst.investigation, activation-watcher-and-archive, sanctions-entry-source, session-case-observability
Cross-Domain Reach: form-entries-and-macro-executions

Investigation Supervisor

Hierarchy Placement: Level 3, child of Case Investigation Desk
Semantic Persona: Routes within the investigation slice of case investigation desk. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: investigation, case, file, note, by, key, value, id, session, search, compile, creation, stage, performance, counter, warning, event, journal
Toolbox (Max 15): none — routes to 3 leaf agents covering 18 tools
Downstream Dependencies: case-record-and-files, case-lookup-and-notes, case-journal-and-events

Case Record and Files (`case-record-and-files`)

Hierarchy Placement: Level 4, child of Investigation Supervisor
Semantic Persona: Owns the case record and files slice end to end, which is small enough that reads and writes stay with one agent. It holds 4 mutating tools, so a confirmation is required before any of them runs.
Semantic Routing Keywords: case, file
Toolbox (Max 15):
- CaseList: Lists active cases for the caller’s tenant, keyset-paged and capped. (permissions from CaseService.RequiredPermissions)
- CaseGet: Returns a single active case by Id, scoped to the caller’s tenant and role. (permissions from CaseService.RequiredPermissions)
- CaseCreateFromCaseKeyValue: Raises a new case (or promotes an existing one) from a matching archived transaction; dispatches the target status’s notification/webhook. Write — confirmation required (permissions from CaseService.RequiredPermissions)
- CaseUpdate: Updates a case’s closed status, lock, diary, rating and workflow status; a status change dispatches that status’s notification/webhook. Write — confirmation required (permissions from CaseService.RequiredPermissions)
- CaseFileRemove: Removes (soft-deletes) a file attached to a Case. Write — confirmation required ( permissions from CaseFileService.RequiredPermissions)
- CaseFileListByCaseKeyValue: Lists the files attached to the Case for a given key field name and value, scoped to the caller’s tenant and Case Workflow roles. (permissions from CaseFileService.RequiredPermissions)
- CaseFileUpload: Uploads and attaches a file to a Case in the caller’s tenant. Write — confirmation required (permissions from CaseFileService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Case Lookup and Notes (`case-lookup-and-notes`)

Hierarchy Placement: Level 4, child of Investigation Supervisor
Semantic Persona: Owns the case lookup and notes slice end to end, which is small enough that reads and writes stay with one agent. It holds 2 mutating tools, so a confirmation is required before any of them runs.
Semantic Routing Keywords: case, note, by, key, value, id, session, search, compile, creation, stage, performance, counter
Toolbox (Max 15):
- CaseNoteInsert: Adds a note to a case, filed against a CaseWorkflowAction; dispatches that action’s notification/webhook if enabled. Write — confirmation required (permissions from CaseNoteService.RequiredPermissions)
- CaseNoteValidate: Validates a Case Note without saving it and returns every failure. (permissions from CaseNoteService.RequiredPermissions)
- CaseNoteListByCaseKeyValue: Lists case notes for a Case Key/Value pair, newest first, scoped to the caller’s tenant. (permissions from CaseNoteService.RequiredPermissions)
- CaseByCaseKeyValueGet: Gets the Cases matching a Case Key and Case Key Value that the caller’s role may see. ( permissions from CaseByCaseKeyValueService.RequiredPermissions)
- CaseByIdGet: Gets one case by Id with its status colours, formatted payload and visible activations; records a case-viewed event. (permissions from CaseByIdService.RequiredPermissions)
- CaseBySessionCaseSearchCompileGet: Retrieves the next case for a compiled Session Case Search, recording a case-opened event and locking the case to the calling user. Write — confirmation required (permissions from CaseBySessionCaseSearchCompileService.RequiredPermissions)
- CaseCreationStagePerformanceCounterList: Lists background case creation stage performance rows – ExistingCasePriorityLookup, WorkflowStatusLookupAndPersist, Notification, HttpEndpoint – most recent first, capped at 100000, with optional date-range and exact-match (StageId, e.g. (permissions from CaseCreationStagePerformanceCounterService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Case Journal and Events (`case-journal-and-events`)

Hierarchy Placement: Level 4, child of Investigation Supervisor
Semantic Persona: Owns the case journal and events slice end to end, which is small enough that reads and writes stay with one agent.
Semantic Routing Keywords: case, creation, warning, event, by, key, value, journal, note
Toolbox (Max 15):
- CaseCreationWarningList: Lists captured case creation warn-threshold breaches – the specific case (workflow, key, key value), the slow stage, and the notification/HTTP destination when relevant – most recent first, capped at 100000, with optional date-range, exac… (permissions from CaseCreationWarningService.RequiredPermissions)
- CaseEventByCaseKeyValueGet: Lists the audit events of a case identified by case key and case key value, newest first. (permissions from CaseEventByCaseKeyValueService.RequiredPermissions)
- CaseJournalGet: Returns the tenant-scoped Case journal of Archive entries matching a payload key and value. ( permissions from CaseJournalService.RequiredPermissions)
- CaseNoteByCaseKeyValueGet: Lists the case notes for cases matching a case key and value, newest first, limited to workflows, statuses and actions the caller’s roles may see within the caller’s tenant. (permissions from CaseNoteByCaseKeyValueService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Activation Watcher and Archive (`activation-watcher-and-archive`)

Hierarchy Placement: Level 3, child of Case Investigation Desk
Semantic Persona: Owns the activation watcher and archive slice end to end, which is small enough that reads and writes stay with one agent. It holds 1 mutating tool, so a confirmation is required before any of them runs.
Semantic Routing Keywords: activation, watcher, archive, archiver, stage, performance, counter, warning
Toolbox (Max 15):
- ActivationWatcherReplay: Replays ActivationWatcher rows created within a date range, scoped to the caller’s tenant. (permissions from ActivationWatcherService.RequiredPermissions)
- ArchiveTag: Replaces the full Tag set on an archived transaction, scoped to the caller’s tenant. Write — confirmation required (permissions from ArchiveService.RequiredPermissions)
- ArchiverStagePerformanceCounterList: Lists background archiver stage performance rows across every Model – BuildArchiveJson, CaseCreationDispatch, RdbmsArchiveWrite, BulkCopyArchiveBuffer – most recent first, capped at 100000, with optional date-range and exact-match (StageI… (permissions from ArchiverStagePerformanceCounterService.RequiredPermissions)
- ArchiverWarningList: Lists captured archiver warn-threshold breaches – the specific invocation (when the stage operates on one item; null for BulkCopyArchiveBuffer), the model, and the slow stage – most recent first, capped at 100000, with optional date-range… (permissions from ArchiverWarningService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Sanctions Entry Source (`sanctions-entry-source`)

Hierarchy Placement: Level 3, child of Case Investigation Desk
Semantic Persona: Owns the sanctions entry source slice end to end, which is small enough that reads and writes stay with one agent.
Semantic Routing Keywords: sanctions, entry, source
Toolbox (Max 15):
- SanctionEntrySourceList: Lists the configured Sanction Entry Sources (e.g. SDN, BOE, EU) the Sanctions Loader can poll or manually import from, each identified by Id and Name. (permissions from SanctionEntrySourceService.RequiredPermissions)
- SanctionEntrySourceFilterFields: Lists the fields a query builder JSON filter over Sanction Entry Sources may use. (permissions from SanctionEntrySourceService.RequiredPermissions)
- SanctionEntrySourceFilter: Returns the Sanction Entry Sources matching query builder JSON, keyset-paged and capped. (permissions from SanctionEntrySourceService.RequiredPermissions)
- SanctionEntrySourceCount: Counts the Sanction Entry Sources matching query builder JSON, optionally grouped by a field. (permissions from SanctionEntrySourceService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Session Case Observability (`session-case-observability`)

Hierarchy Placement: Level 3, child of Case Investigation Desk
Semantic Persona: Owns the session case observability slice end to end, which is small enough that reads and writes stay with one agent. It holds 1 mutating tool, so a confirmation is required before any of them runs.
Semantic Routing Keywords: session, case, journal, search, compiled, sql
Toolbox (Max 15):
- SessionCaseJournalGetByCaseWorkflowGuid: Gets the calling user’s own saved journal (e.g. case grid column layout) for a CaseWorkflow. (permissions from SessionCaseJournalService.RequiredPermissions)
- SessionCaseJournalCreate: Saves the calling user’s own journal for a CaseWorkflow, replacing any earlier save. Write — confirmation required (permissions from SessionCaseJournalService.RequiredPermissions)
- SessionCaseSearchCompiledSqlGetLast: Returns the caller’s own most recently compiled Session Case Search (or a NotFound placeholder); a search flagged for rebuild is recompiled first. (permissions from SessionCaseSearchCompiledSqlService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Case Workflow Governance (`case-workflow-admin`)

Case Workflow Governance

Hierarchy Placement: Level 2, child of Jooby
Semantic Persona: Owns the definitions that shape how cases behave: statuses, routing filters, XPath extraction, forms, actions and macros, and the role grants over each. It designs the workflow an investigator later works inside, and never touches a live case.
Semantic Routing Keywords: case workflow, status, queue, routing, form, macro, action, display, XPath extraction, workflow role
Toolbox (Max 15): none — routes to 7 supervisors covering 121 tools
Downstream Dependencies: case-workflow-admin.definition, case-workflow-admin.entitlement, case-workflow-admin.interaction.actions, case-workflow-admin.interaction.forms, case-workflow-admin.interaction.macros, case-workflow-admin.routing.extraction, case-workflow-admin.routing.status
Cross-Domain Reach: completions

Definition Supervisor

Hierarchy Placement: Level 3, child of Case Workflow Governance
Semantic Persona: Routes within the definition slice of case workflow governance. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: definition, case, workflow
Toolbox (Max 15): none — routes to 3 leaf agents covering 10 tools
Downstream Dependencies: case-workflow-inspector, case-workflow-test-bench, case-workflow-author

Case Workflow Inspector (`case-workflow-inspector`)

Hierarchy Placement: Level 4, child of Definition Supervisor
Semantic Persona: Owns the case workflow slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: case, workflow, inspector
Toolbox (Max 15):
- CaseWorkflowList: Lists Case Workflows visible to the caller’s tenant, keyset-paged and capped. (permissions from CaseWorkflowService.RequiredPermissions)
- CaseWorkflowFilterFields: Lists the fields a query builder JSON filter over Case Workflows may use. ( permissions from CaseWorkflowService.RequiredPermissions)
- CaseWorkflowFilter: Returns the Case Workflows matching query builder JSON, keyset-paged and capped. ( permissions from CaseWorkflowService.RequiredPermissions)
- CaseWorkflowCount: Counts the Case Workflows matching query builder JSON, optionally grouped by a field. ( permissions from CaseWorkflowService.RequiredPermissions)
- CaseWorkflowListByModelActiveOnly: Lists active Case Workflows for a parent model that the caller holds a role grant on. (permissions from CaseWorkflowService.ActiveOnlyPermissions)
- CaseWorkflowGet: Returns a single Case Workflow by Id. (permissions from CaseWorkflowService.RequiredPermissions)
- CaseWorkflowGetByGuid: Returns a single Case Workflow by Id. (permissions from CaseWorkflowService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Case Workflow Test Bench (`case-workflow-test-bench`)

Hierarchy Placement: Level 4, child of Definition Supervisor
Semantic Persona: Owns the case workflow slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: case, workflow, test, bench
Toolbox (Max 15):
- CaseWorkflowValidate: Validates a Case Workflow without saving it and returns every failure. (permissions from CaseWorkflowService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Case Workflow Author (`case-workflow-author`)

Hierarchy Placement: Level 4, child of Definition Supervisor
Semantic Persona: Owns the case workflow slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: case, workflow, author
Toolbox (Max 15):
- CaseWorkflowList: Lists Case Workflows visible to the caller’s tenant, keyset-paged and capped. (permissions from CaseWorkflowService.RequiredPermissions)
- CaseWorkflowFilterFields: Lists the fields a query builder JSON filter over Case Workflows may use. ( permissions from CaseWorkflowService.RequiredPermissions)
- CaseWorkflowFilter: Returns the Case Workflows matching query builder JSON, keyset-paged and capped. ( permissions from CaseWorkflowService.RequiredPermissions)
- CaseWorkflowCount: Counts the Case Workflows matching query builder JSON, optionally grouped by a field. ( permissions from CaseWorkflowService.RequiredPermissions)
- CaseWorkflowListByModelActiveOnly: Lists active Case Workflows for a parent model that the caller holds a role grant on. (permissions from CaseWorkflowService.ActiveOnlyPermissions)
- CaseWorkflowGet: Returns a single Case Workflow by Id. (permissions from CaseWorkflowService.RequiredPermissions)
- CaseWorkflowGetByGuid: Returns a single Case Workflow by Id. (permissions from CaseWorkflowService.RequiredPermissions)
- CaseWorkflowValidate: Validates a Case Workflow without saving it and returns every failure. (permissions from CaseWorkflowService.RequiredPermissions)
- CaseWorkflowCreate: Creates a Case Workflow; calling twice creates two. Write — confirmation required ( permissions from CaseWorkflowService.RequiredPermissions)
- CaseWorkflowUpdate: Updates a Case Workflow; overwrites, writes a version-audit row. Write — confirmation required (permissions from CaseWorkflowService.RequiredPermissions)
- CaseWorkflowDelete: Soft-deletes a Case Workflow; reversible only by direct database action. Write — confirmation required (permissions from CaseWorkflowService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Entitlement Supervisor

Hierarchy Placement: Level 3, child of Case Workflow Governance
Semantic Persona: Routes within the entitlement slice of case workflow governance. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: entitlement, case, workflow, action, role, display, filter, form, macro, status, path
Toolbox (Max 15): none — routes to 4 leaf agents covering 24 tools
Downstream Dependencies: action-and-display-role-grants, filter-and-form-role-grants, macro-and-workflow-role-grants, status-and-xpath-role-grants

Action and Display Role Grants (`action-and-display-role-grants`)

Hierarchy Placement: Level 4, child of Entitlement Supervisor
Semantic Persona: Owns the action and display role grants slice end to end, which is small enough that reads and writes stay with one agent. It holds 4 mutating tools, so a confirmation is required before any of them runs.
Semantic Routing Keywords: case, workflow, action, role, display
Toolbox (Max 15):
- CaseWorkflowActionRoleListByCaseWorkflowActionGuid: Lists the Role grants for a Case Workflow Action. ( permissions from CaseWorkflowActionRoleService.RequiredPermissions)
- CaseWorkflowActionRoleCreate: Grants a Role access to a Case Workflow Action; calling twice creates two grants. Write — confirmation required (permissions from CaseWorkflowActionRoleService.RequiredPermissions)
- CaseWorkflowActionRoleDelete: Removes a Role’s grant on a Case Workflow Action; reversible only by direct database action. Write — confirmation required (permissions from CaseWorkflowActionRoleService.RequiredPermissions)
- CaseWorkflowDisplayRoleListByCaseWorkflowDisplayGuid: Lists the Role grants for a Case Workflow Display. ( permissions from CaseWorkflowDisplayRoleService.RequiredPermissions)
- CaseWorkflowDisplayRoleCreate: Grants a Role access to a Case Workflow Display; calling twice creates two grants. Write — confirmation required (permissions from CaseWorkflowDisplayRoleService.RequiredPermissions)
- CaseWorkflowDisplayRoleDelete: Removes a Role’s grant on a Case Workflow Display; reversible only by direct database action. Write — confirmation required (permissions from CaseWorkflowDisplayRoleService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Filter and Form Role Grants (`filter-and-form-role-grants`)

Hierarchy Placement: Level 4, child of Entitlement Supervisor
Semantic Persona: Owns the filter and form role grants slice end to end, which is small enough that reads and writes stay with one agent. It holds 4 mutating tools, so a confirmation is required before any of them runs.
Semantic Routing Keywords: case, workflow, filter, role, form
Toolbox (Max 15):
- CaseWorkflowFilterRoleListByCaseWorkflowFilterGuid: Lists the Role grants for a Case Workflow Filter. ( permissions from CaseWorkflowFilterRoleService.RequiredPermissions)
- CaseWorkflowFilterRoleCreate: Grants a Role access to a Case Workflow Filter; calling twice creates two grants. Write — confirmation required (permissions from CaseWorkflowFilterRoleService.RequiredPermissions)
- CaseWorkflowFilterRoleDelete: Removes a Role’s grant on a Case Workflow Filter; reversible only by direct database action. Write — confirmation required (permissions from CaseWorkflowFilterRoleService.RequiredPermissions)
- CaseWorkflowFormRoleListByCaseWorkflowFormGuid: Lists the Role grants for a Case Workflow Form. (permissions from CaseWorkflowFormRoleService.RequiredPermissions)
- CaseWorkflowFormRoleCreate: Grants a Role access to a Case Workflow Form; calling twice creates two grants. Write — confirmation required (permissions from CaseWorkflowFormRoleService.RequiredPermissions)
- CaseWorkflowFormRoleDelete: Removes a Role’s grant on a Case Workflow Form; reversible only by direct database action. Write — confirmation required (permissions from CaseWorkflowFormRoleService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Macro and Workflow Role Grants (`macro-and-workflow-role-grants`)

Hierarchy Placement: Level 4, child of Entitlement Supervisor
Semantic Persona: Owns the macro and workflow role grants slice end to end, which is small enough that reads and writes stay with one agent. It holds 4 mutating tools, so a confirmation is required before any of them runs.
Semantic Routing Keywords: case, workflow, macro, role
Toolbox (Max 15):
- CaseWorkflowMacroRoleListByCaseWorkflowMacroGuid: Lists the Role grants for a Case Workflow Macro. (permissions from CaseWorkflowMacroRoleService.RequiredPermissions)
- CaseWorkflowMacroRoleCreate: Grants a Role access to a Case Workflow Macro; calling twice creates two grants. Write — confirmation required (permissions from CaseWorkflowMacroRoleService.RequiredPermissions)
- CaseWorkflowMacroRoleDelete: Removes a Role’s grant on a Case Workflow Macro; reversible only by direct database action. Write — confirmation required (permissions from CaseWorkflowMacroRoleService.RequiredPermissions)
- CaseWorkflowRoleListByCaseWorkflowGuid: Lists the Role grants for a Case Workflow. (permissions from CaseWorkflowRoleService.RequiredPermissions)
- CaseWorkflowRoleCreate: Grants a Role access to a Case Workflow; calling twice creates two grants. Write — confirmation required (permissions from CaseWorkflowRoleService.RequiredPermissions)
- CaseWorkflowRoleDelete: Removes a Role’s grant on a Case Workflow; reversible only by direct database action. Write — confirmation required (permissions from CaseWorkflowRoleService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Status and XPath Role Grants (`status-and-xpath-role-grants`)

Hierarchy Placement: Level 4, child of Entitlement Supervisor
Semantic Persona: Owns the status and xpath role grants slice end to end, which is small enough that reads and writes stay with one agent. It holds 4 mutating tools, so a confirmation is required before any of them runs.
Semantic Routing Keywords: case, workflow, status, role, path
Toolbox (Max 15):
- CaseWorkflowStatusRoleListByCaseWorkflowStatusGuid: Lists the Role grants for a Case Workflow Status. ( permissions from CaseWorkflowStatusRoleService.RequiredPermissions)
- CaseWorkflowStatusRoleCreate: Grants a Role access to a Case Workflow Status; calling twice creates two grants. Write — confirmation required (permissions from CaseWorkflowStatusRoleService.RequiredPermissions)
- CaseWorkflowStatusRoleDelete: Removes a Role’s grant on a Case Workflow Status; reversible only by direct database action. Write — confirmation required (permissions from CaseWorkflowStatusRoleService.RequiredPermissions)
- CaseWorkflowXPathRoleListByCaseWorkflowXPathGuid: Lists the Role grants for a Case Workflow XPath. (permissions from CaseWorkflowXPathRoleService.RequiredPermissions)
- CaseWorkflowXPathRoleCreate: Grants a Role access to a Case Workflow XPath; calling twice creates two grants. Write — confirmation required (permissions from CaseWorkflowXPathRoleService.RequiredPermissions)
- CaseWorkflowXPathRoleDelete: Removes a Role’s grant on a Case Workflow XPath; reversible only by direct database action. Write — confirmation required (permissions from CaseWorkflowXPathRoleService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Interaction Actions Supervisor

Hierarchy Placement: Level 3, child of Case Workflow Governance
Semantic Persona: Routes within the interaction actions slice of case workflow governance. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: interaction actions, case, workflow, action, display
Toolbox (Max 15): none — routes to 6 leaf agents covering 22 tools
Downstream Dependencies: case-workflow-action-inspector, case-workflow-action-test-bench, case-workflow-action-author, case-workflow-display-inspector, case-workflow-display-test-bench, case-workflow-display-author

Case Workflow Action Inspector (`case-workflow-action-inspector`)

Hierarchy Placement: Level 4, child of Interaction Actions Supervisor
Semantic Persona: Owns the case workflow action slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: case, workflow, action, inspector
Toolbox (Max 15):
- CaseWorkflowActionList: Lists Case Workflow Actions visible to the caller’s tenant, keyset-paged and capped. ( permissions from CaseWorkflowActionService.RequiredPermissions)
- CaseWorkflowActionFilterFields: Lists the fields a query builder JSON filter over Case Workflow Actions may use. (permissions from CaseWorkflowActionService.RequiredPermissions)
- CaseWorkflowActionFilter: Returns the Case Workflow Actions matching query builder JSON, keyset-paged and capped. (permissions from CaseWorkflowActionService.RequiredPermissions)
- CaseWorkflowActionCount: Counts the Case Workflow Actions matching query builder JSON, optionally grouped by a field. (permissions from CaseWorkflowActionService.RequiredPermissions)
- CaseWorkflowActionListByWorkflowIdActiveOnly: Lists active Case Workflow Actions for a parent Case Workflow that the caller holds a role grant on. (permissions from CaseWorkflowActionService.ActiveOnlyPermissions)
- CaseWorkflowActionListByWorkflowGuidActiveOnly: Lists active Case Workflow Actions for a parent Case Workflow (by Guid) that the caller holds a role grant on. (permissions from CaseWorkflowActionService.ActiveOnlyPermissions)
- CaseWorkflowActionGet: Returns a single Case Workflow Action by Id. (permissions from CaseWorkflowActionService.RequiredPermissions)
- CaseWorkflowActionGetByGuid: Returns a single Case Workflow Action by Id. (permissions from CaseWorkflowActionService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Case Workflow Action Test Bench (`case-workflow-action-test-bench`)

Hierarchy Placement: Level 4, child of Interaction Actions Supervisor
Semantic Persona: Owns the case workflow action slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: case, workflow, action, test, bench
Toolbox (Max 15):
- CaseWorkflowActionValidate: Validates a Case Workflow Action without saving it and returns every failure. ( permissions from CaseWorkflowActionService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Case Workflow Action Author (`case-workflow-action-author`)

Hierarchy Placement: Level 4, child of Interaction Actions Supervisor
Semantic Persona: Owns the case workflow action slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: case, workflow, action, author
Toolbox (Max 15):
- CaseWorkflowActionList: Lists Case Workflow Actions visible to the caller’s tenant, keyset-paged and capped. ( permissions from CaseWorkflowActionService.RequiredPermissions)
- CaseWorkflowActionFilterFields: Lists the fields a query builder JSON filter over Case Workflow Actions may use. (permissions from CaseWorkflowActionService.RequiredPermissions)
- CaseWorkflowActionFilter: Returns the Case Workflow Actions matching query builder JSON, keyset-paged and capped. (permissions from CaseWorkflowActionService.RequiredPermissions)
- CaseWorkflowActionCount: Counts the Case Workflow Actions matching query builder JSON, optionally grouped by a field. (permissions from CaseWorkflowActionService.RequiredPermissions)
- CaseWorkflowActionListByWorkflowIdActiveOnly: Lists active Case Workflow Actions for a parent Case Workflow that the caller holds a role grant on. (permissions from CaseWorkflowActionService.ActiveOnlyPermissions)
- CaseWorkflowActionListByWorkflowGuidActiveOnly: Lists active Case Workflow Actions for a parent Case Workflow (by Guid) that the caller holds a role grant on. (permissions from CaseWorkflowActionService.ActiveOnlyPermissions)
- CaseWorkflowActionGet: Returns a single Case Workflow Action by Id. (permissions from CaseWorkflowActionService.RequiredPermissions)
- CaseWorkflowActionGetByGuid: Returns a single Case Workflow Action by Id. (permissions from CaseWorkflowActionService.RequiredPermissions)
- CaseWorkflowActionValidate: Validates a Case Workflow Action without saving it and returns every failure. ( permissions from CaseWorkflowActionService.RequiredPermissions)
- CaseWorkflowActionCreate: Creates a Case Workflow Action; calling twice creates two. Write — confirmation required (permissions from CaseWorkflowActionService.RequiredPermissions)
- CaseWorkflowActionUpdate: Updates a Case Workflow Action; overwrites, writes a version-audit row. Write — confirmation required (permissions from CaseWorkflowActionService.RequiredPermissions)
- CaseWorkflowActionDelete: Soft-deletes a Case Workflow Action; reversible only by direct database action. Write — confirmation required (permissions from CaseWorkflowActionService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Case Workflow Display Inspector (`case-workflow-display-inspector`)

Hierarchy Placement: Level 4, child of Interaction Actions Supervisor
Semantic Persona: Owns the case workflow display slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: case, workflow, display, inspector
Toolbox (Max 15):
- CaseWorkflowDisplayList: Lists Case Workflow Displays visible to the caller’s tenant, keyset-paged and capped. (permissions from CaseWorkflowDisplayService.RequiredPermissions)
- CaseWorkflowDisplayFilterFields: Lists the fields a query builder JSON filter over Case Workflow Displays may use. (permissions from CaseWorkflowDisplayService.RequiredPermissions)
- CaseWorkflowDisplayFilter: Returns the Case Workflow Displays matching query builder JSON, keyset-paged and capped. (permissions from CaseWorkflowDisplayService.RequiredPermissions)
- CaseWorkflowDisplayCount: Counts the Case Workflow Displays matching query builder JSON, optionally grouped by a field. (permissions from CaseWorkflowDisplayService.RequiredPermissions)
- CaseWorkflowDisplayListByWorkflowIdActiveOnly: Lists active Case Workflow Displays for a parent Case Workflow that the caller holds a role grant on. (permissions from CaseWorkflowDisplayService.ActiveOnlyPermissions)
- CaseWorkflowDisplayListByWorkflowGuidActiveOnly: Lists active Case Workflow Displays for a parent Case Workflow (by Guid) that the caller holds a role grant on. (permissions from CaseWorkflowDisplayService.ActiveOnlyPermissions)
- CaseWorkflowDisplayGet: Returns a single Case Workflow Display by Id. (permissions from CaseWorkflowDisplayService.RequiredPermissions)
- CaseWorkflowDisplayGetByGuid: Returns a single Case Workflow Display by Id. (permissions from CaseWorkflowDisplayService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Case Workflow Display Test Bench (`case-workflow-display-test-bench`)

Hierarchy Placement: Level 4, child of Interaction Actions Supervisor
Semantic Persona: Owns the case workflow display slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: case, workflow, display, test, bench
Toolbox (Max 15):
- CaseWorkflowDisplayValidate: Validates a Case Workflow Display without saving it and returns every failure. ( permissions from CaseWorkflowDisplayService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Case Workflow Display Author (`case-workflow-display-author`)

Hierarchy Placement: Level 4, child of Interaction Actions Supervisor
Semantic Persona: Owns the case workflow display slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: case, workflow, display, author
Toolbox (Max 15):
- CaseWorkflowDisplayList: Lists Case Workflow Displays visible to the caller’s tenant, keyset-paged and capped. (permissions from CaseWorkflowDisplayService.RequiredPermissions)
- CaseWorkflowDisplayFilterFields: Lists the fields a query builder JSON filter over Case Workflow Displays may use. (permissions from CaseWorkflowDisplayService.RequiredPermissions)
- CaseWorkflowDisplayFilter: Returns the Case Workflow Displays matching query builder JSON, keyset-paged and capped. (permissions from CaseWorkflowDisplayService.RequiredPermissions)
- CaseWorkflowDisplayCount: Counts the Case Workflow Displays matching query builder JSON, optionally grouped by a field. (permissions from CaseWorkflowDisplayService.RequiredPermissions)
- CaseWorkflowDisplayListByWorkflowIdActiveOnly: Lists active Case Workflow Displays for a parent Case Workflow that the caller holds a role grant on. (permissions from CaseWorkflowDisplayService.ActiveOnlyPermissions)
- CaseWorkflowDisplayListByWorkflowGuidActiveOnly: Lists active Case Workflow Displays for a parent Case Workflow (by Guid) that the caller holds a role grant on. (permissions from CaseWorkflowDisplayService.ActiveOnlyPermissions)
- CaseWorkflowDisplayGet: Returns a single Case Workflow Display by Id. (permissions from CaseWorkflowDisplayService.RequiredPermissions)
- CaseWorkflowDisplayGetByGuid: Returns a single Case Workflow Display by Id. (permissions from CaseWorkflowDisplayService.RequiredPermissions)
- CaseWorkflowDisplayValidate: Validates a Case Workflow Display without saving it and returns every failure. ( permissions from CaseWorkflowDisplayService.RequiredPermissions)
- CaseWorkflowDisplayCreate: Creates a Case Workflow Display; calling twice creates two. Write — confirmation required (permissions from CaseWorkflowDisplayService.RequiredPermissions)
- CaseWorkflowDisplayUpdate: Updates a Case Workflow Display; overwrites, writes a version-audit row. Write — confirmation required (permissions from CaseWorkflowDisplayService.RequiredPermissions)
- CaseWorkflowDisplayDelete: Soft-deletes a Case Workflow Display; reversible only by direct database action. Write — confirmation required (permissions from CaseWorkflowDisplayService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Interaction Forms Supervisor

Hierarchy Placement: Level 3, child of Case Workflow Governance
Semantic Persona: Routes within the interaction forms slice of case workflow governance. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: interaction forms, case, workflow, form, entry, display, execution, by, key, value, macro
Toolbox (Max 15): none — routes to 4 leaf agents covering 18 tools
Downstream Dependencies: case-workflow-form-inspector, case-workflow-form-test-bench, case-workflow-form-author, form-entries-and-macro-executions

Case Workflow Form Inspector (`case-workflow-form-inspector`)

Hierarchy Placement: Level 4, child of Interaction Forms Supervisor
Semantic Persona: Owns the case workflow form slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: case, workflow, form, inspector
Toolbox (Max 15):
- CaseWorkflowFormList: Lists Case Workflow Forms visible to the caller’s tenant, keyset-paged and capped. ( permissions from CaseWorkflowFormService.RequiredPermissions)
- CaseWorkflowFormFilterFields: Lists the fields a query builder JSON filter over Case Workflow Forms may use. ( permissions from CaseWorkflowFormService.RequiredPermissions)
- CaseWorkflowFormFilter: Returns the Case Workflow Forms matching query builder JSON, keyset-paged and capped. ( permissions from CaseWorkflowFormService.RequiredPermissions)
- CaseWorkflowFormCount: Counts the Case Workflow Forms matching query builder JSON, optionally grouped by a field. (permissions from CaseWorkflowFormService.RequiredPermissions)
- CaseWorkflowFormListByWorkflowIdActiveOnly: Lists active Case Workflow Forms for a parent Case Workflow that the caller holds a role grant on. (permissions from CaseWorkflowFormService.ActiveOnlyPermissions)
- CaseWorkflowFormListByWorkflowGuidActiveOnly: Lists active Case Workflow Forms for a parent Case Workflow (by Guid) that the caller holds a role grant on. (permissions from CaseWorkflowFormService.ActiveOnlyPermissions)
- CaseWorkflowFormGet: Returns a single Case Workflow Form by Id. (permissions from CaseWorkflowFormService.RequiredPermissions)
- CaseWorkflowFormGetByGuid: Returns a single Case Workflow Form by Id. (permissions from CaseWorkflowFormService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Case Workflow Form Test Bench (`case-workflow-form-test-bench`)

Hierarchy Placement: Level 4, child of Interaction Forms Supervisor
Semantic Persona: Owns the case workflow form slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: case, workflow, form, test, bench
Toolbox (Max 15):
- CaseWorkflowFormValidate: Validates a Case Workflow Form without saving it and returns every failure. ( permissions from CaseWorkflowFormService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Case Workflow Form Author (`case-workflow-form-author`)

Hierarchy Placement: Level 4, child of Interaction Forms Supervisor
Semantic Persona: Owns the case workflow form slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: case, workflow, form, author
Toolbox (Max 15):
- CaseWorkflowFormList: Lists Case Workflow Forms visible to the caller’s tenant, keyset-paged and capped. ( permissions from CaseWorkflowFormService.RequiredPermissions)
- CaseWorkflowFormFilterFields: Lists the fields a query builder JSON filter over Case Workflow Forms may use. ( permissions from CaseWorkflowFormService.RequiredPermissions)
- CaseWorkflowFormFilter: Returns the Case Workflow Forms matching query builder JSON, keyset-paged and capped. ( permissions from CaseWorkflowFormService.RequiredPermissions)
- CaseWorkflowFormCount: Counts the Case Workflow Forms matching query builder JSON, optionally grouped by a field. (permissions from CaseWorkflowFormService.RequiredPermissions)
- CaseWorkflowFormListByWorkflowIdActiveOnly: Lists active Case Workflow Forms for a parent Case Workflow that the caller holds a role grant on. (permissions from CaseWorkflowFormService.ActiveOnlyPermissions)
- CaseWorkflowFormListByWorkflowGuidActiveOnly: Lists active Case Workflow Forms for a parent Case Workflow (by Guid) that the caller holds a role grant on. (permissions from CaseWorkflowFormService.ActiveOnlyPermissions)
- CaseWorkflowFormGet: Returns a single Case Workflow Form by Id. (permissions from CaseWorkflowFormService.RequiredPermissions)
- CaseWorkflowFormGetByGuid: Returns a single Case Workflow Form by Id. (permissions from CaseWorkflowFormService.RequiredPermissions)
- CaseWorkflowFormValidate: Validates a Case Workflow Form without saving it and returns every failure. ( permissions from CaseWorkflowFormService.RequiredPermissions)
- CaseWorkflowFormCreate: Creates a Case Workflow Form; calling twice creates two. Write — confirmation required (permissions from CaseWorkflowFormService.RequiredPermissions)
- CaseWorkflowFormUpdate: Updates a Case Workflow Form; overwrites, writes a version-audit row. Write — confirmation required (permissions from CaseWorkflowFormService.RequiredPermissions)
- CaseWorkflowFormDelete: Soft-deletes a Case Workflow Form; reversible only by direct database action. Write — confirmation required (permissions from CaseWorkflowFormService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Form Entries and Macro Executions (`form-entries-and-macro-executions`)

Hierarchy Placement: Level 4, child of Interaction Forms Supervisor
Semantic Persona: Owns the form entries and macro executions slice end to end, which is small enough that reads and writes stay with one agent. It holds 2 mutating tools, so a confirmation is required before any of them runs.
Semantic Routing Keywords: case, workflow, form, entry, display, execution, by, key, value, macro
Toolbox (Max 15):
- CaseWorkflowFormEntryInsert: Submits a CaseWorkflowForm entry for a case; dispatches that form’s notification/webhook if enabled. Write — confirmation required (permissions from CaseWorkflowFormEntryService.RequiredPermissions)
- CaseWorkflowFormEntryValidate: Validates a Case Workflow Form Entry without saving it and returns every failure. (permissions from CaseWorkflowFormEntryService.RequiredPermissions)
- CaseWorkflowFormEntryListByCaseKeyValue: Lists case workflow form entries for a Case Key/Value pair, newest first, scoped to the caller’s tenant and role grants. (permissions from CaseWorkflowFormEntryService.RequiredPermissions)
- CaseWorkflowDisplayExecutionExecute: Renders a case workflow display HTML template with the case’s stored payload tokens replaced. (permissions from CaseWorkflowDisplayExecutionService.RequiredPermissions)
- CaseWorkflowFormEntryByCaseKeyValueGet: Lists Case Workflow Form Entries for a Case Key and Case Key Value, newest case first, limited to the caller’s tenant and to Case Workflows and Statuses the caller’s role holds. ( permissions from CaseWorkflowFormEntryByCaseKeyValueService.RequiredPermissions)
- CaseWorkflowFormEntryValueListByCaseWorkflowFormEntryId: Lists the submitted field name/value rows for a CaseWorkflowFormEntry, newest first, scoped to the caller’s tenant and role grants. (permissions from CaseWorkflowFormEntryValueService.RequiredPermissions)
- CaseWorkflowMacroExecutionExecute: Executes an active case workflow macro against a case, sending the macro’s configured notification and/or HTTP endpoint call. Write — confirmation required (permissions from CaseWorkflowMacroExecutionService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Interaction Macros Supervisor

Hierarchy Placement: Level 3, child of Case Workflow Governance
Semantic Persona: Routes within the interaction macros slice of case workflow governance. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: interaction macros, case, workflow, macro
Toolbox (Max 15): none — routes to 3 leaf agents covering 11 tools
Downstream Dependencies: case-workflow-macro-inspector, case-workflow-macro-test-bench, case-workflow-macro-author

Case Workflow Macro Inspector (`case-workflow-macro-inspector`)

Hierarchy Placement: Level 4, child of Interaction Macros Supervisor
Semantic Persona: Owns the case workflow macro slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: case, workflow, macro, inspector
Toolbox (Max 15):
- CaseWorkflowMacroList: Lists Case Workflow Macros visible to the caller’s tenant, keyset-paged and capped. ( permissions from CaseWorkflowMacroService.RequiredPermissions)
- CaseWorkflowMacroFilterFields: Lists the fields a query builder JSON filter over Case Workflow Macros may use. (permissions from CaseWorkflowMacroService.RequiredPermissions)
- CaseWorkflowMacroFilter: Returns the Case Workflow Macros matching query builder JSON, keyset-paged and capped. (permissions from CaseWorkflowMacroService.RequiredPermissions)
- CaseWorkflowMacroCount: Counts the Case Workflow Macros matching query builder JSON, optionally grouped by a field. (permissions from CaseWorkflowMacroService.RequiredPermissions)
- CaseWorkflowMacroListByWorkflowIdActiveOnly: Lists active Case Workflow Macros for a parent Case Workflow that the caller holds a role grant on. (permissions from CaseWorkflowMacroService.ActiveOnlyPermissions)
- CaseWorkflowMacroListByWorkflowGuidActiveOnly: Lists active Case Workflow Macros for a parent Case Workflow (by Guid) that the caller holds a role grant on. (permissions from CaseWorkflowMacroService.ActiveOnlyPermissions)
- CaseWorkflowMacroGet: Returns a single Case Workflow Macro by Id. (permissions from CaseWorkflowMacroService.RequiredPermissions)
- CaseWorkflowMacroGetByGuid: Returns a single Case Workflow Macro by Id. (permissions from CaseWorkflowMacroService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Case Workflow Macro Test Bench (`case-workflow-macro-test-bench`)

Hierarchy Placement: Level 4, child of Interaction Macros Supervisor
Semantic Persona: Owns the case workflow macro slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: case, workflow, macro, test, bench
Toolbox (Max 15):
- CaseWorkflowMacroValidate: Validates a Case Workflow Macro without saving it and returns every failure. ( permissions from CaseWorkflowMacroService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Case Workflow Macro Author (`case-workflow-macro-author`)

Hierarchy Placement: Level 4, child of Interaction Macros Supervisor
Semantic Persona: Owns the case workflow macro slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: case, workflow, macro, author
Toolbox (Max 15):
- CaseWorkflowMacroList: Lists Case Workflow Macros visible to the caller’s tenant, keyset-paged and capped. ( permissions from CaseWorkflowMacroService.RequiredPermissions)
- CaseWorkflowMacroFilterFields: Lists the fields a query builder JSON filter over Case Workflow Macros may use. (permissions from CaseWorkflowMacroService.RequiredPermissions)
- CaseWorkflowMacroFilter: Returns the Case Workflow Macros matching query builder JSON, keyset-paged and capped. (permissions from CaseWorkflowMacroService.RequiredPermissions)
- CaseWorkflowMacroCount: Counts the Case Workflow Macros matching query builder JSON, optionally grouped by a field. (permissions from CaseWorkflowMacroService.RequiredPermissions)
- CaseWorkflowMacroListByWorkflowIdActiveOnly: Lists active Case Workflow Macros for a parent Case Workflow that the caller holds a role grant on. (permissions from CaseWorkflowMacroService.ActiveOnlyPermissions)
- CaseWorkflowMacroListByWorkflowGuidActiveOnly: Lists active Case Workflow Macros for a parent Case Workflow (by Guid) that the caller holds a role grant on. (permissions from CaseWorkflowMacroService.ActiveOnlyPermissions)
- CaseWorkflowMacroGet: Returns a single Case Workflow Macro by Id. (permissions from CaseWorkflowMacroService.RequiredPermissions)
- CaseWorkflowMacroGetByGuid: Returns a single Case Workflow Macro by Id. (permissions from CaseWorkflowMacroService.RequiredPermissions)
- CaseWorkflowMacroValidate: Validates a Case Workflow Macro without saving it and returns every failure. ( permissions from CaseWorkflowMacroService.RequiredPermissions)
- CaseWorkflowMacroCreate: Creates a Case Workflow Macro; calling twice creates two. Write — confirmation required (permissions from CaseWorkflowMacroService.RequiredPermissions)
- CaseWorkflowMacroUpdate: Updates a Case Workflow Macro; overwrites, writes a version-audit row. Write — confirmation required (permissions from CaseWorkflowMacroService.RequiredPermissions)
- CaseWorkflowMacroDelete: Soft-deletes a Case Workflow Macro; reversible only by direct database action. Write — confirmation required (permissions from CaseWorkflowMacroService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Routing Extraction Supervisor

Hierarchy Placement: Level 3, child of Case Workflow Governance
Semantic Persona: Routes within the routing extraction slice of case workflow governance. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: routing extraction, case, workflow, filter, path
Toolbox (Max 15): none — routes to 7 leaf agents covering 23 tools
Downstream Dependencies: case-workflow-filter-inspector-1, case-workflow-filter-inspector-2, case-workflow-filter-test-bench, case-workflow-filter-author, case-workflow-x-path-inspector, case-workflow-x-path-test-bench, case-workflow-x-path-author
Cross-Domain Reach: completions

Case Workflow Filter Inspector 1 (`case-workflow-filter-inspector-1`)

Hierarchy Placement: Level 4, child of Routing Extraction Supervisor
Semantic Persona: Owns the case workflow filter slice end to end, which is small enough that reads and writes stay with one agent.
Semantic Routing Keywords: case, workflow, filter, inspector, 1
Toolbox (Max 15):
- CaseWorkflowFilterList: Lists Case Workflow Filters visible to the caller’s tenant, keyset-paged and capped. ( permissions from CaseWorkflowFilterService.ActiveOnlyPermissions)
- CaseWorkflowFilterFilterFields: Lists the fields a query builder JSON filter over Case Workflow Filters may use. (permissions from CaseWorkflowFilterService.ActiveOnlyPermissions)
- CaseWorkflowFilterFilter: Returns the Case Workflow Filters matching query builder JSON, keyset-paged and capped. (permissions from CaseWorkflowFilterService.ActiveOnlyPermissions)
- CaseWorkflowFilterCount: Counts the Case Workflow Filters matching query builder JSON, optionally grouped by a field. (permissions from CaseWorkflowFilterService.ActiveOnlyPermissions)
- CaseWorkflowFilterListByWorkflowIdActiveOnly: Lists active Case Workflow Filters for a parent Case Workflow that the caller holds a role grant on. (permissions from CaseWorkflowFilterService.ActiveOnlyPermissions)
- CaseWorkflowFilterListByWorkflowGuidActiveOnly: Lists active Case Workflow Filters for a parent Case Workflow (by Guid) that the caller holds a role grant on. (permissions from CaseWorkflowFilterService.ActiveOnlyPermissions)
- CaseWorkflowFilterGet: Returns a single Case Workflow Filter by Id. (permissions from CaseWorkflowFilterService.ActiveOnlyPermissions)
Downstream Dependencies: none (leaf)

Case Workflow Filter Inspector 2 (`case-workflow-filter-inspector-2`)

Hierarchy Placement: Level 4, child of Routing Extraction Supervisor
Semantic Persona: Owns the case workflow filter slice end to end, which is small enough that reads and writes stay with one agent.
Semantic Routing Keywords: case, workflow, filter, inspector, 2
Toolbox (Max 15):
- CaseWorkflowFilterGetByGuid: Returns a single Case Workflow Filter by Guid. (permissions from CaseWorkflowFilterService.ActiveOnlyPermissions)
Downstream Dependencies: none (leaf)

Case Workflow Filter Test Bench (`case-workflow-filter-test-bench`)

Hierarchy Placement: Level 4, child of Routing Extraction Supervisor
Semantic Persona: Owns the case workflow filter slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: case, workflow, filter, test, bench
Toolbox (Max 15):
- CaseWorkflowFilterValidate: Validates a Case Workflow Filter without saving it and returns every failure. ( permissions from CaseWorkflowFilterService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Case Workflow Filter Author (`case-workflow-filter-author`)

Hierarchy Placement: Level 4, child of Routing Extraction Supervisor
Semantic Persona: Owns the case workflow filter slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: case, workflow, filter, author
Toolbox (Max 15):
- CaseWorkflowFilterList: Lists Case Workflow Filters visible to the caller’s tenant, keyset-paged and capped. ( permissions from CaseWorkflowFilterService.ActiveOnlyPermissions)
- CaseWorkflowFilterFilterFields: Lists the fields a query builder JSON filter over Case Workflow Filters may use. (permissions from CaseWorkflowFilterService.ActiveOnlyPermissions)
- CaseWorkflowFilterFilter: Returns the Case Workflow Filters matching query builder JSON, keyset-paged and capped. (permissions from CaseWorkflowFilterService.ActiveOnlyPermissions)
- CaseWorkflowFilterCount: Counts the Case Workflow Filters matching query builder JSON, optionally grouped by a field. (permissions from CaseWorkflowFilterService.ActiveOnlyPermissions)
- CaseWorkflowFilterListByWorkflowIdActiveOnly: Lists active Case Workflow Filters for a parent Case Workflow that the caller holds a role grant on. (permissions from CaseWorkflowFilterService.ActiveOnlyPermissions)
- CaseWorkflowFilterListByWorkflowGuidActiveOnly: Lists active Case Workflow Filters for a parent Case Workflow (by Guid) that the caller holds a role grant on. (permissions from CaseWorkflowFilterService.ActiveOnlyPermissions)
- CaseWorkflowFilterGet: Returns a single Case Workflow Filter by Id. (permissions from CaseWorkflowFilterService.ActiveOnlyPermissions)
- CaseWorkflowFilterGetByGuid: Returns a single Case Workflow Filter by Guid. (permissions from CaseWorkflowFilterService.ActiveOnlyPermissions)
- CaseWorkflowFilterValidate: Validates a Case Workflow Filter without saving it and returns every failure. ( permissions from CaseWorkflowFilterService.RequiredPermissions)
- CaseWorkflowFilterCreate: Creates a Case Workflow Filter; calling twice creates two. Write — confirmation required (permissions from CaseWorkflowFilterService.RequiredPermissions)
- CaseWorkflowFilterUpdate: Updates a Case Workflow Filter; overwrites, writes a version-audit row. Write — confirmation required (permissions from CaseWorkflowFilterService.RequiredPermissions)
- CaseWorkflowFilterDelete: Soft-deletes a Case Workflow Filter; reversible only by direct database action. Write — confirmation required (permissions from CaseWorkflowFilterService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Case Workflow X Path Inspector (`case-workflow-x-path-inspector`)

Hierarchy Placement: Level 4, child of Routing Extraction Supervisor
Semantic Persona: Owns the case workflow x path slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: case, workflow, path, inspector
Toolbox (Max 15):
- CaseWorkflowXPathList: Lists Case Workflow XPath extractions visible to the caller’s tenant, keyset-paged and capped. (permissions from CaseWorkflowXPathService.RequiredPermissions)
- CaseWorkflowXPathFilterFields: Lists the fields a query builder JSON filter over Case WorkflowXPaths may use. ( permissions from CaseWorkflowXPathService.RequiredPermissions)
- CaseWorkflowXPathFilter: Returns the Case WorkflowXPaths matching query builder JSON, keyset-paged and capped. (permissions from CaseWorkflowXPathService.RequiredPermissions)
- CaseWorkflowXPathCount: Counts the Case WorkflowXPaths matching query builder JSON, optionally grouped by a field. (permissions from CaseWorkflowXPathService.RequiredPermissions)
- CaseWorkflowXPathListByWorkflowIdActiveDrillOnly: Lists active, Drill-enabled Case Workflow XPath extractions for a parent Case Workflow that the caller holds a role grant on. (permissions from CaseWorkflowXPathService.ActiveOnlyPermissions)
- CaseWorkflowXPathListByWorkflowGuidActiveDrillOnly: Lists active, Drill-enabled Case Workflow XPath extractions for a parent Case Workflow (by Guid) that the caller holds a role grant on. (permissions from CaseWorkflowXPathService.ActiveOnlyPermissions)
- CaseWorkflowXPathGet: Returns a single Case Workflow XPath extraction by Id. (permissions from CaseWorkflowXPathService.RequiredPermissions)
- CaseWorkflowXPathGetByGuid: Returns a single Case Workflow XPath extraction by Id. (permissions from CaseWorkflowXPathService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Case Workflow X Path Test Bench (`case-workflow-x-path-test-bench`)

Hierarchy Placement: Level 4, child of Routing Extraction Supervisor
Semantic Persona: Owns the case workflow x path slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: case, workflow, path, test, bench
Toolbox (Max 15):
- CaseWorkflowXPathValidate: Validates a Case WorkflowXPath without saving it and returns every failure. ( permissions from CaseWorkflowXPathService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Case Workflow X Path Author (`case-workflow-x-path-author`)

Hierarchy Placement: Level 4, child of Routing Extraction Supervisor
Semantic Persona: Owns the case workflow x path slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: case, workflow, path, author
Toolbox (Max 15):
- CaseWorkflowXPathList: Lists Case Workflow XPath extractions visible to the caller’s tenant, keyset-paged and capped. (permissions from CaseWorkflowXPathService.RequiredPermissions)
- CaseWorkflowXPathFilterFields: Lists the fields a query builder JSON filter over Case WorkflowXPaths may use. ( permissions from CaseWorkflowXPathService.RequiredPermissions)
- CaseWorkflowXPathFilter: Returns the Case WorkflowXPaths matching query builder JSON, keyset-paged and capped. (permissions from CaseWorkflowXPathService.RequiredPermissions)
- CaseWorkflowXPathCount: Counts the Case WorkflowXPaths matching query builder JSON, optionally grouped by a field. (permissions from CaseWorkflowXPathService.RequiredPermissions)
- CaseWorkflowXPathListByWorkflowIdActiveDrillOnly: Lists active, Drill-enabled Case Workflow XPath extractions for a parent Case Workflow that the caller holds a role grant on. (permissions from CaseWorkflowXPathService.ActiveOnlyPermissions)
- CaseWorkflowXPathListByWorkflowGuidActiveDrillOnly: Lists active, Drill-enabled Case Workflow XPath extractions for a parent Case Workflow (by Guid) that the caller holds a role grant on. (permissions from CaseWorkflowXPathService.ActiveOnlyPermissions)
- CaseWorkflowXPathGet: Returns a single Case Workflow XPath extraction by Id. (permissions from CaseWorkflowXPathService.RequiredPermissions)
- CaseWorkflowXPathGetByGuid: Returns a single Case Workflow XPath extraction by Id. (permissions from CaseWorkflowXPathService.RequiredPermissions)
- CaseWorkflowXPathValidate: Validates a Case WorkflowXPath without saving it and returns every failure. ( permissions from CaseWorkflowXPathService.RequiredPermissions)
- CaseWorkflowXPathCreate: Creates a Case Workflow XPath extraction; calling twice creates two. Write — confirmation required (permissions from CaseWorkflowXPathService.RequiredPermissions)
- CaseWorkflowXPathUpdate: Updates a Case Workflow XPath extraction; overwrites, writes a version-audit row. Write — confirmation required (permissions from CaseWorkflowXPathService.RequiredPermissions)
- CaseWorkflowXPathDelete: Soft-deletes a Case Workflow XPath extraction; reversible only by direct database action. Write — confirmation required (permissions from CaseWorkflowXPathService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Routing Status Supervisor

Hierarchy Placement: Level 3, child of Case Workflow Governance
Semantic Persona: Routes within the routing status slice of case workflow governance. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: routing status, case, workflow, status, priority
Toolbox (Max 15): none — routes to 5 leaf agents covering 13 tools
Downstream Dependencies: case-workflow-status-inspector-1, case-workflow-status-inspector-2, case-workflow-status-test-bench, case-workflow-status-author, case-workflow-priority

Case Workflow Status Inspector 1 (`case-workflow-status-inspector-1`)

Hierarchy Placement: Level 4, child of Routing Status Supervisor
Semantic Persona: Owns the case workflow status slice end to end, which is small enough that reads and writes stay with one agent.
Semantic Routing Keywords: case, workflow, status, inspector, 1
Toolbox (Max 15):
- CaseWorkflowStatusList: Lists Case Workflow Statuses visible to the caller’s tenant, keyset-paged and capped. ( permissions from CaseWorkflowStatusService.RequiredPermissions)
- CaseWorkflowStatusFilterFields: Lists the fields a query builder JSON filter over Case Workflow Statuses may use. (permissions from CaseWorkflowStatusService.RequiredPermissions)
- CaseWorkflowStatusFilter: Returns the Case Workflow Statuses matching query builder JSON, keyset-paged and capped. (permissions from CaseWorkflowStatusService.RequiredPermissions)
- CaseWorkflowStatusCount: Counts the Case Workflow Statuses matching query builder JSON, optionally grouped by a field. (permissions from CaseWorkflowStatusService.RequiredPermissions)
- CaseWorkflowStatusListByWorkflowGuid: Lists Case Workflow Statuses for a parent Case Workflow (by Guid), including inactive ones. (permissions from CaseWorkflowStatusService.ListByWorkflowPermissions)
- CaseWorkflowStatusListByWorkflowIdActiveOnly: Lists active Case Workflow Statuses for a parent Case Workflow that the caller holds a role grant on. (permissions from CaseWorkflowStatusService.ActiveOnlyPermissions)
- CaseWorkflowStatusListByWorkflowGuidActiveOnly: Lists active Case Workflow Statuses for a parent Case Workflow (by Guid) that the caller holds a role grant on. (permissions from CaseWorkflowStatusService.ActiveOnlyPermissions)
Downstream Dependencies: none (leaf)

Case Workflow Status Inspector 2 (`case-workflow-status-inspector-2`)

Hierarchy Placement: Level 4, child of Routing Status Supervisor
Semantic Persona: Owns the case workflow status slice end to end, which is small enough that reads and writes stay with one agent.
Semantic Routing Keywords: case, workflow, status, inspector, 2
Toolbox (Max 15):
- CaseWorkflowStatusGet: Returns a single Case Workflow Status by Id. (permissions from CaseWorkflowStatusService.RequiredPermissions)
- CaseWorkflowStatusGetByGuid: Returns a single Case Workflow Status by Id. (permissions from CaseWorkflowStatusService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Case Workflow Status Test Bench (`case-workflow-status-test-bench`)

Hierarchy Placement: Level 4, child of Routing Status Supervisor
Semantic Persona: Owns the case workflow status slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: case, workflow, status, test, bench
Toolbox (Max 15):
- CaseWorkflowStatusValidate: Validates a Case Workflow Status without saving it and returns every failure. ( permissions from CaseWorkflowStatusService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Case Workflow Status Author (`case-workflow-status-author`)

Hierarchy Placement: Level 4, child of Routing Status Supervisor
Semantic Persona: Owns the case workflow status slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: case, workflow, status, author
Toolbox (Max 15):
- CaseWorkflowStatusList: Lists Case Workflow Statuses visible to the caller’s tenant, keyset-paged and capped. ( permissions from CaseWorkflowStatusService.RequiredPermissions)
- CaseWorkflowStatusFilterFields: Lists the fields a query builder JSON filter over Case Workflow Statuses may use. (permissions from CaseWorkflowStatusService.RequiredPermissions)
- CaseWorkflowStatusFilter: Returns the Case Workflow Statuses matching query builder JSON, keyset-paged and capped. (permissions from CaseWorkflowStatusService.RequiredPermissions)
- CaseWorkflowStatusCount: Counts the Case Workflow Statuses matching query builder JSON, optionally grouped by a field. (permissions from CaseWorkflowStatusService.RequiredPermissions)
- CaseWorkflowStatusListByWorkflowGuid: Lists Case Workflow Statuses for a parent Case Workflow (by Guid), including inactive ones. (permissions from CaseWorkflowStatusService.ListByWorkflowPermissions)
- CaseWorkflowStatusListByWorkflowIdActiveOnly: Lists active Case Workflow Statuses for a parent Case Workflow that the caller holds a role grant on. (permissions from CaseWorkflowStatusService.ActiveOnlyPermissions)
- CaseWorkflowStatusListByWorkflowGuidActiveOnly: Lists active Case Workflow Statuses for a parent Case Workflow (by Guid) that the caller holds a role grant on. (permissions from CaseWorkflowStatusService.ActiveOnlyPermissions)
- CaseWorkflowStatusGet: Returns a single Case Workflow Status by Id. (permissions from CaseWorkflowStatusService.RequiredPermissions)
- CaseWorkflowStatusGetByGuid: Returns a single Case Workflow Status by Id. (permissions from CaseWorkflowStatusService.RequiredPermissions)
- CaseWorkflowStatusValidate: Validates a Case Workflow Status without saving it and returns every failure. ( permissions from CaseWorkflowStatusService.RequiredPermissions)
- CaseWorkflowStatusCreate: Creates a Case Workflow Status; calling twice creates two. Write — confirmation required (permissions from CaseWorkflowStatusService.RequiredPermissions)
- CaseWorkflowStatusUpdate: Updates a Case Workflow Status; overwrites, writes a version-audit row. Write — confirmation required (permissions from CaseWorkflowStatusService.RequiredPermissions)
- CaseWorkflowStatusDelete: Soft-deletes a Case Workflow Status; reversible only by direct database action. Write — confirmation required (permissions from CaseWorkflowStatusService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Case Workflow Priority (`case-workflow-priority`)

Hierarchy Placement: Level 4, child of Routing Status Supervisor
Semantic Persona: Owns the case workflow priority slice end to end, which is small enough that reads and writes stay with one agent.
Semantic Routing Keywords: case, workflow, priority
Toolbox (Max 15):
- CaseWorkflowPriorityList: Lists the fixed Case priority levels (High, Medium, Low) – the same for every Case Workflow and every tenant. (permissions from CaseWorkflowPriorityService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Analytics Reporting (`data-analyst`)

Analytics Reporting

Hierarchy Placement: Level 2, child of Jooby
Semantic Persona: Owns the visualisation registry, its datasources, parameters and the role grants over them. Produces and governs the reporting surface, and reads detection output without ever altering it.
Semantic Routing Keywords: visualisation, dashboard, chart, report, datasource, parameter, series, directory
Toolbox (Max 15): none — routes to 3 supervisors covering 44 tools
Downstream Dependencies: data-analyst.visualisation.definition, data-analyst.visualisation.parameters, data-analyst.visualisation.registry

Visualisation Definition Supervisor

Hierarchy Placement: Level 3, child of Analytics Reporting
Semantic Persona: Routes within the visualisation definition slice of analytics reporting. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: visualisation definition, visualisation, registry, role, datasource, command, execution, series
Toolbox (Max 15): none — routes to 4 leaf agents covering 16 tools
Downstream Dependencies: visualisation-registry-inspector, visualisation-registry-test-bench, visualisation-registry-author, visualisation-role-grants-and-execution

Visualisation Registry Inspector (`visualisation-registry-inspector`)

Hierarchy Placement: Level 4, child of Visualisation Definition Supervisor
Semantic Persona: Owns the visualisation registry slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: visualisation, registry, inspector
Toolbox (Max 15):
- VisualisationRegistryList: Lists Visualisation Registries visible to the caller’s tenant, keyset-paged and capped. (permissions from VisualisationRegistryService.ListPermissions)
- VisualisationRegistryFilterFields: Lists the fields a query builder JSON filter over Visualisation Registrys may use. (permissions from VisualisationRegistryService.ListPermissions)
- VisualisationRegistryFilter: Returns the Visualisation Registrys matching query builder JSON, keyset-paged and capped. (permissions from VisualisationRegistryService.ListPermissions)
- VisualisationRegistryCount: Counts the Visualisation Registrys matching query builder JSON, optionally grouped by a field. (permissions from VisualisationRegistryService.ListPermissions)
- VisualisationRegistryGet: Returns a single Visualisation Registry by Id. (permissions from VisualisationRegistryService.ReadPermissions)
- VisualisationRegistryGetByGuid: Returns a single Visualisation Registry by Id. (permissions from VisualisationRegistryService.ReadPermissions)
- VisualisationRegistryGetByGuidActiveOnly: Returns a single active Visualisation Registry by Guid, for embedded recall, provided the caller holds a role grant on it. (permissions from VisualisationRegistryService.ReadPermissions)
- VisualisationRegistryListByShowInDirectory: Lists active, Show In Directory Visualisation Registries that the caller holds a role grant on. (permissions from VisualisationRegistryService.ShowInDirectoryPermissions)
Downstream Dependencies: none (leaf)

Visualisation Registry Test Bench (`visualisation-registry-test-bench`)

Hierarchy Placement: Level 4, child of Visualisation Definition Supervisor
Semantic Persona: Owns the visualisation registry slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: visualisation, registry, test, bench
Toolbox (Max 15):
- VisualisationRegistryValidate: Validates a Visualisation Registry without saving it and returns every failure. (permissions from VisualisationRegistryService.WritePermissions)
Downstream Dependencies: none (leaf)

Visualisation Registry Author (`visualisation-registry-author`)

Hierarchy Placement: Level 4, child of Visualisation Definition Supervisor
Semantic Persona: Owns the visualisation registry slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: visualisation, registry, author
Toolbox (Max 15):
- VisualisationRegistryList: Lists Visualisation Registries visible to the caller’s tenant, keyset-paged and capped. (permissions from VisualisationRegistryService.ListPermissions)
- VisualisationRegistryFilterFields: Lists the fields a query builder JSON filter over Visualisation Registrys may use. (permissions from VisualisationRegistryService.ListPermissions)
- VisualisationRegistryFilter: Returns the Visualisation Registrys matching query builder JSON, keyset-paged and capped. (permissions from VisualisationRegistryService.ListPermissions)
- VisualisationRegistryCount: Counts the Visualisation Registrys matching query builder JSON, optionally grouped by a field. (permissions from VisualisationRegistryService.ListPermissions)
- VisualisationRegistryGet: Returns a single Visualisation Registry by Id. (permissions from VisualisationRegistryService.ReadPermissions)
- VisualisationRegistryGetByGuid: Returns a single Visualisation Registry by Id. (permissions from VisualisationRegistryService.ReadPermissions)
- VisualisationRegistryGetByGuidActiveOnly: Returns a single active Visualisation Registry by Guid, for embedded recall, provided the caller holds a role grant on it. (permissions from VisualisationRegistryService.ReadPermissions)
- VisualisationRegistryListByShowInDirectory: Lists active, Show In Directory Visualisation Registries that the caller holds a role grant on. (permissions from VisualisationRegistryService.ShowInDirectoryPermissions)
- VisualisationRegistryValidate: Validates a Visualisation Registry without saving it and returns every failure. (permissions from VisualisationRegistryService.WritePermissions)
- VisualisationRegistryCreate: Creates a Visualisation Registry; calling twice creates two. Write — confirmation required (permissions from VisualisationRegistryService.WritePermissions)
- VisualisationRegistryUpdate: Updates a Visualisation Registry; overwrites, writes a version-audit row. Write — confirmation required (permissions from VisualisationRegistryService.WritePermissions)
- VisualisationRegistryDelete: Soft-deletes a Visualisation Registry; reversible only by direct database action. Write — confirmation required (permissions from VisualisationRegistryService.WritePermissions)
Downstream Dependencies: none (leaf)

Visualisation Role Grants and Execution (`visualisation-role-grants-and-execution`)

Hierarchy Placement: Level 4, child of Visualisation Definition Supervisor
Semantic Persona: Owns the visualisation role grants and execution slice end to end, which is small enough that reads and writes stay with one agent. It holds 2 mutating tools, so a confirmation is required before any of them runs.
Semantic Routing Keywords: visualisation, registry, role, datasource, command, execution, series
Toolbox (Max 15):
- VisualisationRegistryRoleListByVisualisationRegistryGuid: Lists the Role grants for a Visualisation Registry. ( permissions from VisualisationRegistryRoleService.RequiredPermissions)
- VisualisationRegistryRoleCreate: Grants a Role access to a Visualisation Registry; calling twice creates two grants. Write — confirmation required (permissions from VisualisationRegistryRoleService.RequiredPermissions)
- VisualisationRegistryRoleDelete: Removes a Role’s grant on a Visualisation Registry; reversible only by direct database action. Write — confirmation required (permissions from VisualisationRegistryRoleService.RequiredPermissions)
- VisualisationRegistryDatasourceCommandExecutionExecute: Executes a Visualisation Registry Datasource command with caller-supplied parameter values and returns its rows. (permissions from VisualisationRegistryDatasourceCommandExecutionService.RequiredPermissions)
- VisualisationRegistryDatasourceSeriesListByVisualisationRegistryDatasourceId: Lists the series definitions (name and inferred data type of each SQL result column) of a Visualisation Registry Datasource; server-derived, never user-created. (permissions from VisualisationRegistryDatasourceSeriesService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Visualisation Parameters Supervisor

Hierarchy Placement: Level 3, child of Analytics Reporting
Semantic Persona: Routes within the visualisation parameters slice of analytics reporting. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: visualisation parameters, visualisation, registry, parameter
Toolbox (Max 15): none — routes to 3 leaf agents covering 11 tools
Downstream Dependencies: visualisation-registry-parameter-inspector, visualisation-registry-parameter-test-bench, visualisation-registry-parameter-author

Visualisation Registry Parameter Inspector (`visualisation-registry-parameter-inspector`)

Hierarchy Placement: Level 4, child of Visualisation Parameters Supervisor
Semantic Persona: Owns the visualisation registry parameter slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: visualisation, registry, parameter, inspector
Toolbox (Max 15):
- VisualisationRegistryParameterList: Lists Visualisation Registry Parameters for the caller’s tenant, keyset-paged and capped. (permissions from VisualisationRegistryParameterService.RequiredPermissions)
- VisualisationRegistryParameterFilterFields: Lists the fields a query builder JSON filter over Visualisation Registry Parameters may use. (permissions from VisualisationRegistryParameterService.RequiredPermissions)
- VisualisationRegistryParameterFilter: Returns the Visualisation Registry Parameters matching query builder JSON, keyset-paged and capped. (permissions from VisualisationRegistryParameterService.RequiredPermissions)
- VisualisationRegistryParameterCount: Counts the Visualisation Registry Parameters matching query builder JSON, optionally grouped by a field. (permissions from VisualisationRegistryParameterService.RequiredPermissions)
- VisualisationRegistryParameterGet: Returns one Visualisation Registry Parameter by its id, scoped to the caller’s tenant. (permissions from VisualisationRegistryParameterService.RequiredPermissions)
- VisualisationRegistryParameterGetByGuid: Returns one Visualisation Registry Parameter by its id, scoped to the caller’s tenant. (permissions from VisualisationRegistryParameterService.RequiredPermissions)
- VisualisationRegistryParameterGetByVisualisationRegistryId: Lists the Parameters belonging to a given Visualisation Registry, ordered by Id, scoped to the caller’s tenant. (permissions from VisualisationRegistryParameterService.ActiveOnlyPermissions)
- VisualisationRegistryParameterGetByVisualisationRegistryIdActiveOnly: Lists the active Parameters belonging to a given Visualisation Registry that the caller holds a role grant on, scoped to the caller’s tenant. (permissions from VisualisationRegistryParameterService.ActiveOnlyPermissions)
Downstream Dependencies: none (leaf)

Visualisation Registry Parameter Test Bench (`visualisation-registry-parameter-test-bench`)

Hierarchy Placement: Level 4, child of Visualisation Parameters Supervisor
Semantic Persona: Owns the visualisation registry parameter slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: visualisation, registry, parameter, test, bench
Toolbox (Max 15):
- VisualisationRegistryParameterValidate: Validates a Visualisation Registry Parameter without saving it and returns every failure. (permissions from VisualisationRegistryParameterService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Visualisation Registry Parameter Author (`visualisation-registry-parameter-author`)

Hierarchy Placement: Level 4, child of Visualisation Parameters Supervisor
Semantic Persona: Owns the visualisation registry parameter slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: visualisation, registry, parameter, author
Toolbox (Max 15):
- VisualisationRegistryParameterList: Lists Visualisation Registry Parameters for the caller’s tenant, keyset-paged and capped. (permissions from VisualisationRegistryParameterService.RequiredPermissions)
- VisualisationRegistryParameterFilterFields: Lists the fields a query builder JSON filter over Visualisation Registry Parameters may use. (permissions from VisualisationRegistryParameterService.RequiredPermissions)
- VisualisationRegistryParameterFilter: Returns the Visualisation Registry Parameters matching query builder JSON, keyset-paged and capped. (permissions from VisualisationRegistryParameterService.RequiredPermissions)
- VisualisationRegistryParameterCount: Counts the Visualisation Registry Parameters matching query builder JSON, optionally grouped by a field. (permissions from VisualisationRegistryParameterService.RequiredPermissions)
- VisualisationRegistryParameterGet: Returns one Visualisation Registry Parameter by its id, scoped to the caller’s tenant. (permissions from VisualisationRegistryParameterService.RequiredPermissions)
- VisualisationRegistryParameterGetByGuid: Returns one Visualisation Registry Parameter by its id, scoped to the caller’s tenant. (permissions from VisualisationRegistryParameterService.RequiredPermissions)
- VisualisationRegistryParameterGetByVisualisationRegistryId: Lists the Parameters belonging to a given Visualisation Registry, ordered by Id, scoped to the caller’s tenant. (permissions from VisualisationRegistryParameterService.ActiveOnlyPermissions)
- VisualisationRegistryParameterGetByVisualisationRegistryIdActiveOnly: Lists the active Parameters belonging to a given Visualisation Registry that the caller holds a role grant on, scoped to the caller’s tenant. (permissions from VisualisationRegistryParameterService.ActiveOnlyPermissions)
- VisualisationRegistryParameterValidate: Validates a Visualisation Registry Parameter without saving it and returns every failure. (permissions from VisualisationRegistryParameterService.RequiredPermissions)
- VisualisationRegistryParameterCreate: Registers a Parameter under a Visualisation Registry in the caller’s tenant; calling twice creates two. Write — confirmation required (permissions from VisualisationRegistryParameterService.RequiredPermissions)
- VisualisationRegistryParameterUpdate: Updates a Parameter in the caller’s tenant by id. Write — confirmation required (permissions from VisualisationRegistryParameterService.RequiredPermissions)
- VisualisationRegistryParameterDelete: Soft-deletes a Parameter in the caller’s tenant by id. Write — confirmation required (permissions from VisualisationRegistryParameterService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Visualisation Registry Supervisor

Hierarchy Placement: Level 3, child of Analytics Reporting
Semantic Persona: Routes within the visualisation registry slice of analytics reporting. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: visualisation registry, visualisation, registry, datasource, role, parameter
Toolbox (Max 15): none — routes to 4 leaf agents covering 17 tools
Downstream Dependencies: visualisation-registry-datasource-inspector, visualisation-registry-datasource-test-bench, visualisation-registry-datasource-author, datasource-and-parameter-role-grants
Cross-Domain Reach: visualisation-registry-inspector, visualisation-registry-author, visualisation-registry-parameter-inspector, visualisation-registry-parameter-author

Visualisation Registry Datasource Inspector (`visualisation-registry-datasource-inspector`)

Hierarchy Placement: Level 4, child of Visualisation Registry Supervisor
Semantic Persona: Owns the visualisation registry datasource slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: visualisation, registry, datasource, inspector
Toolbox (Max 15):
- VisualisationRegistryDatasourceList: Lists Datasources for the caller’s tenant, keyset-paged and capped. ( permissions from VisualisationRegistryDatasourceService.ListPermissions)
- VisualisationRegistryDatasourceFilterFields: Lists the fields a query builder JSON filter over Visualisation Registry Datasources may use. (permissions from VisualisationRegistryDatasourceService.ListPermissions)
- VisualisationRegistryDatasourceFilter: Returns the Visualisation Registry Datasources matching query builder JSON, keyset-paged and capped. (permissions from VisualisationRegistryDatasourceService.ListPermissions)
- VisualisationRegistryDatasourceCount: Counts the Visualisation Registry Datasources matching query builder JSON, optionally grouped by a field. (permissions from VisualisationRegistryDatasourceService.ListPermissions)
- VisualisationRegistryDatasourceGet: Returns one Datasource by id, scoped to the caller’s tenant. (permissions from VisualisationRegistryDatasourceService.ReadPermissions)
- VisualisationRegistryDatasourceGetByGuid: Returns one Datasource by Guid, scoped to the caller’s tenant. ( permissions from VisualisationRegistryDatasourceService.ReadPermissions)
- VisualisationRegistryDatasourceGetByVisualisationRegistryId: Lists Datasources belonging to a given Visualisation Registry, scoped to the caller’s tenant. (permissions from VisualisationRegistryDatasourceService.ReadByParentPermissions)
- VisualisationRegistryDatasourceGetByVisualisationRegistryIdActiveOnly: Lists active, role-visible Datasources belonging to a given Visualisation Registry for the caller. (permissions from VisualisationRegistryDatasourceService.ReadByParentPermissions)
Downstream Dependencies: none (leaf)

Visualisation Registry Datasource Test Bench (`visualisation-registry-datasource-test-bench`)

Hierarchy Placement: Level 4, child of Visualisation Registry Supervisor
Semantic Persona: Owns the visualisation registry datasource slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: visualisation, registry, datasource, test, bench
Toolbox (Max 15):
- VisualisationRegistryDatasourceValidate: Validates a Visualisation Registry Datasource without saving it and returns every failure. (permissions from VisualisationRegistryDatasourceService.WritePermissions)
Downstream Dependencies: none (leaf)

Visualisation Registry Datasource Author (`visualisation-registry-datasource-author`)

Hierarchy Placement: Level 4, child of Visualisation Registry Supervisor
Semantic Persona: Owns the visualisation registry datasource slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: visualisation, registry, datasource, author
Toolbox (Max 15):
- VisualisationRegistryDatasourceList: Lists Datasources for the caller’s tenant, keyset-paged and capped. ( permissions from VisualisationRegistryDatasourceService.ListPermissions)
- VisualisationRegistryDatasourceFilterFields: Lists the fields a query builder JSON filter over Visualisation Registry Datasources may use. (permissions from VisualisationRegistryDatasourceService.ListPermissions)
- VisualisationRegistryDatasourceFilter: Returns the Visualisation Registry Datasources matching query builder JSON, keyset-paged and capped. (permissions from VisualisationRegistryDatasourceService.ListPermissions)
- VisualisationRegistryDatasourceCount: Counts the Visualisation Registry Datasources matching query builder JSON, optionally grouped by a field. (permissions from VisualisationRegistryDatasourceService.ListPermissions)
- VisualisationRegistryDatasourceGet: Returns one Datasource by id, scoped to the caller’s tenant. (permissions from VisualisationRegistryDatasourceService.ReadPermissions)
- VisualisationRegistryDatasourceGetByGuid: Returns one Datasource by Guid, scoped to the caller’s tenant. ( permissions from VisualisationRegistryDatasourceService.ReadPermissions)
- VisualisationRegistryDatasourceGetByVisualisationRegistryId: Lists Datasources belonging to a given Visualisation Registry, scoped to the caller’s tenant. (permissions from VisualisationRegistryDatasourceService.ReadByParentPermissions)
- VisualisationRegistryDatasourceGetByVisualisationRegistryIdActiveOnly: Lists active, role-visible Datasources belonging to a given Visualisation Registry for the caller. (permissions from VisualisationRegistryDatasourceService.ReadByParentPermissions)
- VisualisationRegistryDatasourceValidate: Validates a Visualisation Registry Datasource without saving it and returns every failure. (permissions from VisualisationRegistryDatasourceService.WritePermissions)
- VisualisationRegistryDatasourceCreate: Registers a Datasource under a Visualisation Registry in the caller’s tenant, validating its SQL Command; calling twice creates two. Write — confirmation required (permissions from VisualisationRegistryDatasourceService.WritePermissions)
- VisualisationRegistryDatasourceUpdate: Updates a Datasource in the caller’s tenant by id, re-validating its SQL Command. Write — confirmation required (permissions from VisualisationRegistryDatasourceService.WritePermissions)
- VisualisationRegistryDatasourceDelete: Soft-deletes a Datasource in the caller’s tenant by id. Write — confirmation required (permissions from VisualisationRegistryDatasourceService.WritePermissions)
Downstream Dependencies: none (leaf)

Datasource and Parameter Role Grants (`datasource-and-parameter-role-grants`)

Hierarchy Placement: Level 4, child of Visualisation Registry Supervisor
Semantic Persona: Owns the datasource and parameter role grants slice end to end, which is small enough that reads and writes stay with one agent. It holds 4 mutating tools, so a confirmation is required before any of them runs.
Semantic Routing Keywords: visualisation, registry, datasource, role, parameter
Toolbox (Max 15):
- VisualisationRegistryDatasourceRoleListByVisualisationRegistryDatasourceGuid: Lists the Role grants for a Visualisation Registry Datasource. (permissions from VisualisationRegistryDatasourceRoleService.RequiredPermissions)
- VisualisationRegistryDatasourceRoleCreate: Grants a Role access to a Visualisation Registry Datasource; calling twice creates two grants. Write — confirmation required (permissions from VisualisationRegistryDatasourceRoleService.RequiredPermissions)
- VisualisationRegistryDatasourceRoleDelete: Removes a Role’s grant on a Visualisation Registry Datasource; reversible only by direct database action. Write — confirmation required (permissions from VisualisationRegistryDatasourceRoleService.RequiredPermissions)
- VisualisationRegistryParameterRoleListByVisualisationRegistryParameterGuid: Lists the Role grants for a Visualisation Registry Parameter. (permissions from VisualisationRegistryParameterRoleService.RequiredPermissions)
- VisualisationRegistryParameterRoleCreate: Grants a Role access to a Visualisation Registry Parameter; calling twice creates two grants. Write — confirmation required (permissions from VisualisationRegistryParameterRoleService.RequiredPermissions)
- VisualisationRegistryParameterRoleDelete: Removes a Role’s grant on a Visualisation Registry Parameter; reversible only by direct database action. Write — confirmation required (permissions from VisualisationRegistryParameterRoleService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Detection Model Authoring (`model-admin`)

Detection Model Authoring

Hierarchy Placement: Level 2, child of Jooby
Semantic Persona: Authors the detection model itself and the reference and abstraction material it is built from: request payload contract, derivations, gateway and activation rules, counters and aggregations. This is where fraud and AML typologies become executable rules.
Semantic Routing Keywords: model, rule, activation, gateway, abstraction, TTL counter, request XPath, inline function, inline script, tag, sanction list, threshold
Toolbox (Max 15): none — routes to 8 supervisors covering 153 tools
Downstream Dependencies: model-admin.abstraction.calculations, model-admin.abstraction.counters, model-admin.activation, model-admin.core, model-admin.references.derivation, model-admin.references.gateway, model-admin.references.payload-contract, model-admin.references.watchlists
Cross-Domain Reach: completions

Abstraction Calculations Supervisor

Hierarchy Placement: Level 3, child of Detection Model Authoring
Semantic Persona: Routes within the abstraction calculations slice of detection model authoring. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: abstraction calculations, entity, analysis, model, abstraction, calculation, rule
Toolbox (Max 15): none — routes to 6 leaf agents covering 25 tools
Downstream Dependencies: entity-analysis-model-abstraction-calculation-inspector, entity-analysis-model-abstraction-calculation-test-bench, entity-analysis-model-abstraction-calculation-author, entity-analysis-model-abstraction-rule-inspector, entity-analysis-model-abstraction-rule-test-bench, entity-analysis-model-abstraction-rule-author
Cross-Domain Reach: completions, model-invocation-context-and-dependencies

Entity Analysis Model Abstraction Calculation Inspector (`entity-analysis-model-abstraction-calculation-inspector`)

Hierarchy Placement: Level 4, child of Abstraction Calculations Supervisor
Semantic Persona: Owns the entity analysis model abstraction calculation slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: entity, analysis, model, abstraction, calculation, inspector
Toolbox (Max 15):
- EntityAnalysisModelAbstractionCalculationList: Lists Abstraction Calculations for the caller’s tenant, keyset-paged and capped. (permissions from EntityAnalysisModelAbstractionCalculationService.ListPermissions)
- EntityAnalysisModelAbstractionCalculationFilterFields: Lists the fields a query builder JSON filter over Abstraction Calculations may use. (permissions from EntityAnalysisModelAbstractionCalculationService.ListPermissions)
- EntityAnalysisModelAbstractionCalculationFilter: Returns the Abstraction Calculations matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelAbstractionCalculationService.ListPermissions)
- EntityAnalysisModelAbstractionCalculationCount: Counts the Abstraction Calculations matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelAbstractionCalculationService.ListPermissions)
- EntityAnalysisModelAbstractionCalculationGet: Returns one Abstraction Calculation by id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelAbstractionCalculationService.ReadPermissions)
- EntityAnalysisModelAbstractionCalculationGetByGuid: Returns one Abstraction Calculation by Guid, scoped to the caller’s tenant. (permissions from EntityAnalysisModelAbstractionCalculationService.ReadPermissions)
- EntityAnalysisModelAbstractionCalculationGetByEntityAnalysisModelId: Lists Abstraction Calculations belonging to a given Model, scoped to the caller’s tenant. (permissions from EntityAnalysisModelAbstractionCalculationService.ReadPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Abstraction Calculation Test Bench (`entity-analysis-model-abstraction-calculation-test-bench`)

Hierarchy Placement: Level 4, child of Abstraction Calculations Supervisor
Semantic Persona: Owns the entity analysis model abstraction calculation slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: entity, analysis, model, abstraction, calculation, test, bench
Toolbox (Max 15):
- EntityAnalysisModelAbstractionCalculationValidate: Validates an Abstraction Calculation without saving it and returns every failure. (permissions from EntityAnalysisModelAbstractionCalculationService.WritePermissions)
- EntityAnalysisModelAbstractionCalculationParseRule: Parses and compiles an Abstraction Calculation’s rule text as the engine would and returns each located error. (permissions from EntityAnalysisModelAbstractionCalculationService.WritePermissions)
- EntityAnalysisModelAbstractionCalculationExecute: Runs an Abstraction Calculation against an invocation context as the engine would and returns the result. (permissions from EntityAnalysisModelAbstractionCalculationService.WritePermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Abstraction Calculation Author (`entity-analysis-model-abstraction-calculation-author`)

Hierarchy Placement: Level 4, child of Abstraction Calculations Supervisor
Semantic Persona: Owns the entity analysis model abstraction calculation slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: entity, analysis, model, abstraction, calculation, author
Toolbox (Max 15):
- EntityAnalysisModelAbstractionCalculationList: Lists Abstraction Calculations for the caller’s tenant, keyset-paged and capped. (permissions from EntityAnalysisModelAbstractionCalculationService.ListPermissions)
- EntityAnalysisModelAbstractionCalculationFilterFields: Lists the fields a query builder JSON filter over Abstraction Calculations may use. (permissions from EntityAnalysisModelAbstractionCalculationService.ListPermissions)
- EntityAnalysisModelAbstractionCalculationFilter: Returns the Abstraction Calculations matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelAbstractionCalculationService.ListPermissions)
- EntityAnalysisModelAbstractionCalculationCount: Counts the Abstraction Calculations matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelAbstractionCalculationService.ListPermissions)
- EntityAnalysisModelAbstractionCalculationGet: Returns one Abstraction Calculation by id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelAbstractionCalculationService.ReadPermissions)
- EntityAnalysisModelAbstractionCalculationGetByGuid: Returns one Abstraction Calculation by Guid, scoped to the caller’s tenant. (permissions from EntityAnalysisModelAbstractionCalculationService.ReadPermissions)
- EntityAnalysisModelAbstractionCalculationGetByEntityAnalysisModelId: Lists Abstraction Calculations belonging to a given Model, scoped to the caller’s tenant. (permissions from EntityAnalysisModelAbstractionCalculationService.ReadPermissions)
- EntityAnalysisModelAbstractionCalculationValidate: Validates an Abstraction Calculation without saving it and returns every failure. (permissions from EntityAnalysisModelAbstractionCalculationService.WritePermissions)
- EntityAnalysisModelAbstractionCalculationParseRule: Parses and compiles an Abstraction Calculation’s rule text as the engine would and returns each located error. (permissions from EntityAnalysisModelAbstractionCalculationService.WritePermissions)
- EntityAnalysisModelAbstractionCalculationExecute: Runs an Abstraction Calculation against an invocation context as the engine would and returns the result. (permissions from EntityAnalysisModelAbstractionCalculationService.WritePermissions)
- EntityAnalysisModelAbstractionCalculationCreate: Registers an Abstraction Calculation under a Model in the caller’s tenant; calling twice creates two. Write — confirmation required (permissions from EntityAnalysisModelAbstractionCalculationService.WritePermissions)
- EntityAnalysisModelAbstractionCalculationUpdate: Updates an Abstraction Calculation in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelAbstractionCalculationService.WritePermissions)
- EntityAnalysisModelAbstractionCalculationDelete: Soft-deletes an Abstraction Calculation in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelAbstractionCalculationService.WritePermissions)
- EntityAnalysisModelAbstractionCalculationResubmit: Resubmits a rejected Abstraction Calculation for approval again, unchanged, opening a fresh approval cycle. Write — confirmation required (permissions from EntityAnalysisModelAbstractionCalculationService.WritePermissions)
- EntityAnalysisModelAbstractionCalculationRetract: Retracts a rejected Abstraction Calculation, restoring it to its last approved version without a new approval cycle. Write — confirmation required (permissions from EntityAnalysisModelAbstractionCalculationService.WritePermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Abstraction Rule Inspector (`entity-analysis-model-abstraction-rule-inspector`)

Hierarchy Placement: Level 4, child of Abstraction Calculations Supervisor
Semantic Persona: Owns the entity analysis model abstraction rule slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: entity, analysis, model, abstraction, rule, inspector
Toolbox (Max 15):
- EntityAnalysisModelAbstractionRuleList: Lists Abstraction Rules for the caller’s tenant, keyset-paged and capped. (permissions from EntityAnalysisModelAbstractionRuleService.ListPermissions)
- EntityAnalysisModelAbstractionRuleFilterFields: Lists the fields a query builder JSON filter over Abstraction Rules may use. (permissions from EntityAnalysisModelAbstractionRuleService.ListPermissions)
- EntityAnalysisModelAbstractionRuleFilter: Returns the Abstraction Rules matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelAbstractionRuleService.ListPermissions)
- EntityAnalysisModelAbstractionRuleCount: Counts the Abstraction Rules matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelAbstractionRuleService.ListPermissions)
- EntityAnalysisModelAbstractionRuleGet: Returns one Abstraction Rule by id, scoped to the caller’s tenant. ( permissions from EntityAnalysisModelAbstractionRuleService.ReadPermissions)
- EntityAnalysisModelAbstractionRuleGetByGuid: Returns one Abstraction Rule by Guid, scoped to the caller’s tenant. (permissions from EntityAnalysisModelAbstractionRuleService.ReadPermissions)
- EntityAnalysisModelAbstractionRuleGetByEntityAnalysisModelId: Lists Abstraction Rules belonging to a given Model, scoped to the caller’s tenant. (permissions from EntityAnalysisModelAbstractionRuleService.ReadPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Abstraction Rule Test Bench (`entity-analysis-model-abstraction-rule-test-bench`)

Hierarchy Placement: Level 4, child of Abstraction Calculations Supervisor
Semantic Persona: Owns the entity analysis model abstraction rule slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: entity, analysis, model, abstraction, rule, test, bench
Toolbox (Max 15):
- EntityAnalysisModelAbstractionRuleValidate: Validates an Abstraction Rule without saving it and returns every failure. (permissions from EntityAnalysisModelAbstractionRuleService.WritePermissions)
- EntityAnalysisModelAbstractionRuleParseRule: Parses and compiles an Abstraction Rule’s rule text as the engine would and returns each located error. (permissions from EntityAnalysisModelAbstractionRuleService.WritePermissions)
- EntityAnalysisModelAbstractionRuleExecute: Runs an Abstraction Rule against an invocation context as the engine would and returns the result. (permissions from EntityAnalysisModelAbstractionRuleService.WritePermissions)
- EntityAnalysisModelAbstractionRuleBuildRuleFromBuilderJson: Turns query builder JSON into an Abstraction Rule’s rule text as the browser builder would, and parses it. (permissions from EntityAnalysisModelAbstractionRuleService.WritePermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Abstraction Rule Author (`entity-analysis-model-abstraction-rule-author`)

Hierarchy Placement: Level 4, child of Abstraction Calculations Supervisor
Semantic Persona: Owns the entity analysis model abstraction rule slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: entity, analysis, model, abstraction, rule, author
Toolbox (Max 15):
- EntityAnalysisModelAbstractionRuleList: Lists Abstraction Rules for the caller’s tenant, keyset-paged and capped. (permissions from EntityAnalysisModelAbstractionRuleService.ListPermissions)
- EntityAnalysisModelAbstractionRuleFilterFields: Lists the fields a query builder JSON filter over Abstraction Rules may use. (permissions from EntityAnalysisModelAbstractionRuleService.ListPermissions)
- EntityAnalysisModelAbstractionRuleFilter: Returns the Abstraction Rules matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelAbstractionRuleService.ListPermissions)
- EntityAnalysisModelAbstractionRuleCount: Counts the Abstraction Rules matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelAbstractionRuleService.ListPermissions)
- EntityAnalysisModelAbstractionRuleGet: Returns one Abstraction Rule by id, scoped to the caller’s tenant. ( permissions from EntityAnalysisModelAbstractionRuleService.ReadPermissions)
- EntityAnalysisModelAbstractionRuleGetByGuid: Returns one Abstraction Rule by Guid, scoped to the caller’s tenant. (permissions from EntityAnalysisModelAbstractionRuleService.ReadPermissions)
- EntityAnalysisModelAbstractionRuleGetByEntityAnalysisModelId: Lists Abstraction Rules belonging to a given Model, scoped to the caller’s tenant. (permissions from EntityAnalysisModelAbstractionRuleService.ReadPermissions)
- EntityAnalysisModelAbstractionRuleValidate: Validates an Abstraction Rule without saving it and returns every failure. (permissions from EntityAnalysisModelAbstractionRuleService.WritePermissions)
- EntityAnalysisModelAbstractionRuleParseRule: Parses and compiles an Abstraction Rule’s rule text as the engine would and returns each located error. (permissions from EntityAnalysisModelAbstractionRuleService.WritePermissions)
- EntityAnalysisModelAbstractionRuleExecute: Runs an Abstraction Rule against an invocation context as the engine would and returns the result. (permissions from EntityAnalysisModelAbstractionRuleService.WritePermissions)
- EntityAnalysisModelAbstractionRuleBuildRuleFromBuilderJson: Turns query builder JSON into an Abstraction Rule’s rule text as the browser builder would, and parses it. (permissions from EntityAnalysisModelAbstractionRuleService.WritePermissions)
- EntityAnalysisModelAbstractionRuleCreate: Registers an Abstraction Rule under a Model in the caller’s tenant; calling twice creates two. Write — confirmation required (permissions from EntityAnalysisModelAbstractionRuleService.WritePermissions)
- EntityAnalysisModelAbstractionRuleUpdate: Updates an Abstraction Rule in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelAbstractionRuleService.WritePermissions)
- EntityAnalysisModelAbstractionRuleDelete: Soft-deletes an Abstraction Rule in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelAbstractionRuleService.WritePermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Abstraction Rule Approvals (`entity-analysis-model-abstraction-rule-approvals`)

Hierarchy Placement: Level 4, child of Abstraction Calculations Supervisor
Semantic Persona: Owns the maker-checker aftermath of the entity analysis model abstraction rule slice: resolves a rejected submission by resubmitting it unchanged or retracting it to its last approved version. Never creates, updates or deletes it.
Semantic Routing Keywords: entity, analysis, model, abstraction, rule, approvals, resubmit, retract
Toolbox (Max 15):
- EntityAnalysisModelAbstractionRuleResubmit: Resubmits a rejected Abstraction Rule for approval again, unchanged, opening a fresh approval cycle. Write — confirmation required (permissions from EntityAnalysisModelAbstractionRuleService.WritePermissions)
- EntityAnalysisModelAbstractionRuleRetract: Retracts a rejected Abstraction Rule, restoring it to its last approved version without a new approval cycle. Write — confirmation required (permissions from EntityAnalysisModelAbstractionRuleService.WritePermissions)
Downstream Dependencies: none (leaf)

Abstraction Counters Supervisor

Hierarchy Placement: Level 3, child of Detection Model Authoring
Semantic Persona: Routes within the abstraction counters slice of detection model authoring. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: abstraction counters, entity, analysis, model, ttl, counter, time, window
Toolbox (Max 15): none — routes to 4 leaf agents covering 15 tools
Downstream Dependencies: entity-analysis-model-ttl-counter-inspector, entity-analysis-model-ttl-counter-test-bench, entity-analysis-model-ttl-counter-author, entity-analysis-model-time-window
Cross-Domain Reach: completions

Entity Analysis Model Ttl Counter Inspector (`entity-analysis-model-ttl-counter-inspector`)

Hierarchy Placement: Level 4, child of Abstraction Counters Supervisor
Semantic Persona: Owns the entity analysis model ttl counter slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: entity, analysis, model, ttl, counter, inspector
Toolbox (Max 15):
- EntityAnalysisModelTtlCounterList: Lists TTL Counters for the caller’s tenant, keyset-paged and capped. ( permissions from EntityAnalysisModelTtlCounterService.ListPermissions)
- EntityAnalysisModelTtlCounterFilterFields: Lists the fields a query builder JSON filter over TTL Counters may use. (permissions from EntityAnalysisModelTtlCounterService.ListPermissions)
- EntityAnalysisModelTtlCounterFilter: Returns the TTL Counters matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelTtlCounterService.ListPermissions)
- EntityAnalysisModelTtlCounterCount: Counts the TTL Counters matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelTtlCounterService.ListPermissions)
- EntityAnalysisModelTtlCounterGet: Returns one TTL Counter by id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelTtlCounterService.ReadPermissions)
- EntityAnalysisModelTtlCounterGetByGuid: Returns one TTL Counter by Guid, scoped to the caller’s tenant. ( permissions from EntityAnalysisModelTtlCounterService.ReadPermissions)
- EntityAnalysisModelTtlCounterGetByEntityAnalysisModelId: Lists TTL Counters belonging to a given Model, scoped to the caller’s tenant. (permissions from EntityAnalysisModelTtlCounterService.ReadByModelPermissions)
- EntityAnalysisModelTtlCounterGetByEntityAnalysisModelGuid: Lists TTL Counters belonging to the Model with the given Guid, scoped to the caller’s tenant. (permissions from EntityAnalysisModelTtlCounterService.ReadByModelPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Ttl Counter Test Bench (`entity-analysis-model-ttl-counter-test-bench`)

Hierarchy Placement: Level 4, child of Abstraction Counters Supervisor
Semantic Persona: Owns the entity analysis model ttl counter slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: entity, analysis, model, ttl, counter, test, bench
Toolbox (Max 15):
- EntityAnalysisModelTtlCounterValidate: Validates a TTL Counter without saving it and returns every failure. ( permissions from EntityAnalysisModelTtlCounterService.WritePermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Ttl Counter Author (`entity-analysis-model-ttl-counter-author`)

Hierarchy Placement: Level 4, child of Abstraction Counters Supervisor
Semantic Persona: Owns the entity analysis model ttl counter slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: entity, analysis, model, ttl, counter, author
Toolbox (Max 15):
- EntityAnalysisModelTtlCounterList: Lists TTL Counters for the caller’s tenant, keyset-paged and capped. ( permissions from EntityAnalysisModelTtlCounterService.ListPermissions)
- EntityAnalysisModelTtlCounterFilterFields: Lists the fields a query builder JSON filter over TTL Counters may use. (permissions from EntityAnalysisModelTtlCounterService.ListPermissions)
- EntityAnalysisModelTtlCounterFilter: Returns the TTL Counters matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelTtlCounterService.ListPermissions)
- EntityAnalysisModelTtlCounterCount: Counts the TTL Counters matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelTtlCounterService.ListPermissions)
- EntityAnalysisModelTtlCounterGet: Returns one TTL Counter by id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelTtlCounterService.ReadPermissions)
- EntityAnalysisModelTtlCounterGetByGuid: Returns one TTL Counter by Guid, scoped to the caller’s tenant. ( permissions from EntityAnalysisModelTtlCounterService.ReadPermissions)
- EntityAnalysisModelTtlCounterGetByEntityAnalysisModelId: Lists TTL Counters belonging to a given Model, scoped to the caller’s tenant. (permissions from EntityAnalysisModelTtlCounterService.ReadByModelPermissions)
- EntityAnalysisModelTtlCounterGetByEntityAnalysisModelGuid: Lists TTL Counters belonging to the Model with the given Guid, scoped to the caller’s tenant. (permissions from EntityAnalysisModelTtlCounterService.ReadByModelPermissions)
- EntityAnalysisModelTtlCounterValidate: Validates a TTL Counter without saving it and returns every failure. ( permissions from EntityAnalysisModelTtlCounterService.WritePermissions)
- EntityAnalysisModelTtlCounterCreate: Registers a TTL Counter under a Model in the caller’s tenant; calling twice creates two. Write — confirmation required (permissions from EntityAnalysisModelTtlCounterService.WritePermissions)
- EntityAnalysisModelTtlCounterUpdate: Updates a TTL Counter in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelTtlCounterService.WritePermissions)
- EntityAnalysisModelTtlCounterDelete: Soft-deletes a TTL Counter in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelTtlCounterService.WritePermissions)
- EntityAnalysisModelTtlCounterResubmit: Resubmits a rejected TTL Counter for approval again, unchanged, opening a fresh approval cycle. Write — confirmation required (permissions from EntityAnalysisModelTtlCounterService.WritePermissions)
- EntityAnalysisModelTtlCounterRetract: Retracts a rejected TTL Counter, restoring it to its last approved version without a new approval cycle. Write — confirmation required (permissions from EntityAnalysisModelTtlCounterService.WritePermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Time Window (`entity-analysis-model-time-window`)

Hierarchy Placement: Level 4, child of Abstraction Counters Supervisor
Semantic Persona: Owns the entity analysis model time window slice end to end, which is small enough that reads and writes stay with one agent.
Semantic Routing Keywords: entity, analysis, model, time, window
Toolbox (Max 15):
- EntityAnalysisModelTimeWindowIntervals: Lists the kinds of time window and the interval codes each allows. ( permissions from EntityAnalysisModelTimeWindowService.RequiredPermissions)
- EntityAnalysisModelTimeWindowCalculate: Works out a time window’s exact start and end as the engine does. ( permissions from EntityAnalysisModelTimeWindowService.RequiredPermissions)
- EntityAnalysisModelTimeWindowAbstractionRule: Works out the window an abstraction rule aggregates over for a reference date. (permissions from EntityAnalysisModelTimeWindowService.RequiredPermissions)
- EntityAnalysisModelTimeWindowTtlCounter: Works out the window a TTL counter counts over for a reference date. ( permissions from EntityAnalysisModelTimeWindowService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Activation Supervisor

Hierarchy Placement: Level 3, child of Detection Model Authoring
Semantic Persona: Routes within the activation slice of detection model authoring. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: activation, entity, analysis, model, rule, override, query
Toolbox (Max 15): none — routes to 7 leaf agents covering 27 tools
Downstream Dependencies: entity-analysis-model-activation-rule-inspector, entity-analysis-model-activation-rule-test-bench, entity-analysis-model-activation-rule-author, entity-analysis-model-activation-rule-override-inspector, entity-analysis-model-activation-rule-override-test-bench, entity-analysis-model-activation-rule-override-author, entity-analysis-model-activation-rule-override-query
Cross-Domain Reach: completions, model-invocation-context-and-dependencies

Entity Analysis Model Activation Rule Inspector (`entity-analysis-model-activation-rule-inspector`)

Hierarchy Placement: Level 4, child of Activation Supervisor
Semantic Persona: Owns the entity analysis model activation rule slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: entity, analysis, model, activation, rule, inspector
Toolbox (Max 15):
- EntityAnalysisModelActivationRuleList: Lists Activation Rules for the caller’s tenant, keyset-paged and capped. (permissions from EntityAnalysisModelActivationRuleService.ListPermissions)
- EntityAnalysisModelActivationRuleFilterFields: Lists the fields a query builder JSON filter over Activation Rules may use. (permissions from EntityAnalysisModelActivationRuleService.ListPermissions)
- EntityAnalysisModelActivationRuleFilter: Returns the Activation Rules matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelActivationRuleService.ListPermissions)
- EntityAnalysisModelActivationRuleCount: Counts the Activation Rules matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelActivationRuleService.ListPermissions)
- EntityAnalysisModelActivationRuleGet: Returns one Activation Rule by id, scoped to the caller’s tenant. ( permissions from EntityAnalysisModelActivationRuleService.ReadPermissions)
- EntityAnalysisModelActivationRuleGetByGuid: Returns one Activation Rule by Guid, scoped to the caller’s tenant. (permissions from EntityAnalysisModelActivationRuleService.ReadPermissions)
- EntityAnalysisModelActivationRuleGetByEntityAnalysisModelId: Lists Activation Rules belonging to a given Model, scoped to the caller’s tenant. (permissions from EntityAnalysisModelActivationRuleService.ReadPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Activation Rule Test Bench (`entity-analysis-model-activation-rule-test-bench`)

Hierarchy Placement: Level 4, child of Activation Supervisor
Semantic Persona: Owns the entity analysis model activation rule slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: entity, analysis, model, activation, rule, test, bench
Toolbox (Max 15):
- EntityAnalysisModelActivationRuleValidate: Validates an Activation Rule without saving it and returns every failure. (permissions from EntityAnalysisModelActivationRuleService.WritePermissions)
- EntityAnalysisModelActivationRuleParseRule: Parses and compiles an Activation Rule’s rule text as the engine would and returns each located error. (permissions from EntityAnalysisModelActivationRuleService.WritePermissions)
- EntityAnalysisModelActivationRuleExecute: Runs an Activation Rule against an invocation context as the engine would and returns the result. (permissions from EntityAnalysisModelActivationRuleService.WritePermissions)
- EntityAnalysisModelActivationRuleBuildRuleFromBuilderJson: Turns query builder JSON into an Activation Rule’s rule text as the browser builder would, and parses it. (permissions from EntityAnalysisModelActivationRuleService.WritePermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Activation Rule Author (`entity-analysis-model-activation-rule-author`)

Hierarchy Placement: Level 4, child of Activation Supervisor
Semantic Persona: Owns the entity analysis model activation rule slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: entity, analysis, model, activation, rule, author
Toolbox (Max 15):
- EntityAnalysisModelActivationRuleList: Lists Activation Rules for the caller’s tenant, keyset-paged and capped. (permissions from EntityAnalysisModelActivationRuleService.ListPermissions)
- EntityAnalysisModelActivationRuleFilterFields: Lists the fields a query builder JSON filter over Activation Rules may use. (permissions from EntityAnalysisModelActivationRuleService.ListPermissions)
- EntityAnalysisModelActivationRuleFilter: Returns the Activation Rules matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelActivationRuleService.ListPermissions)
- EntityAnalysisModelActivationRuleCount: Counts the Activation Rules matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelActivationRuleService.ListPermissions)
- EntityAnalysisModelActivationRuleGet: Returns one Activation Rule by id, scoped to the caller’s tenant. ( permissions from EntityAnalysisModelActivationRuleService.ReadPermissions)
- EntityAnalysisModelActivationRuleGetByGuid: Returns one Activation Rule by Guid, scoped to the caller’s tenant. (permissions from EntityAnalysisModelActivationRuleService.ReadPermissions)
- EntityAnalysisModelActivationRuleGetByEntityAnalysisModelId: Lists Activation Rules belonging to a given Model, scoped to the caller’s tenant. (permissions from EntityAnalysisModelActivationRuleService.ReadPermissions)
- EntityAnalysisModelActivationRuleValidate: Validates an Activation Rule without saving it and returns every failure. (permissions from EntityAnalysisModelActivationRuleService.WritePermissions)
- EntityAnalysisModelActivationRuleParseRule: Parses and compiles an Activation Rule’s rule text as the engine would and returns each located error. (permissions from EntityAnalysisModelActivationRuleService.WritePermissions)
- EntityAnalysisModelActivationRuleExecute: Runs an Activation Rule against an invocation context as the engine would and returns the result. (permissions from EntityAnalysisModelActivationRuleService.WritePermissions)
- EntityAnalysisModelActivationRuleBuildRuleFromBuilderJson: Turns query builder JSON into an Activation Rule’s rule text as the browser builder would, and parses it. (permissions from EntityAnalysisModelActivationRuleService.WritePermissions)
- EntityAnalysisModelActivationRuleCreate: Registers an Activation Rule under a Model in the caller’s tenant; calling twice creates two. Write — confirmation required (permissions from EntityAnalysisModelActivationRuleService.WritePermissions)
- EntityAnalysisModelActivationRuleUpdate: Updates an Activation Rule in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelActivationRuleService.WritePermissions)
- EntityAnalysisModelActivationRuleDelete: Soft-deletes an Activation Rule in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelActivationRuleService.WritePermissions)
- EntityAnalysisModelActivationRuleReset: Resets the ActivationCounter and EvaluationCounter of an Activation Rule to zero. Write — confirmation required (permissions from EntityAnalysisModelActivationRuleService.WritePermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Activation Rule Approvals (`entity-analysis-model-activation-rule-approvals`)

Hierarchy Placement: Level 4, child of Core Supervisor (the Activation Supervisor’s own fan-out was already at budget)
Semantic Persona: Owns the maker-checker aftermath of the entity analysis model activation rule slice: resolves a rejected submission by resubmitting it unchanged or retracting it to its last approved version. Never creates, updates or deletes it.
Semantic Routing Keywords: entity, analysis, model, activation, rule, approvals, resubmit, retract
Toolbox (Max 15):
- EntityAnalysisModelActivationRuleResubmit: Resubmits a rejected Activation Rule for approval again, unchanged, opening a fresh approval cycle. Write — confirmation required (permissions from EntityAnalysisModelActivationRuleService.WritePermissions)
- EntityAnalysisModelActivationRuleRetract: Retracts a rejected Activation Rule, restoring it to its last approved version without a new approval cycle. Write — confirmation required (permissions from EntityAnalysisModelActivationRuleService.WritePermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Activation Rule Override Inspector (`entity-analysis-model-activation-rule-override-inspector`)

Hierarchy Placement: Level 4, child of Activation Supervisor
Semantic Persona: Owns the entity analysis model activation rule override slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: entity, analysis, model, activation, rule, override, inspector
Toolbox (Max 15):
- EntityAnalysisModelActivationRuleOverrideList: Lists Activation-Rule-scoped Overrides for the caller’s tenant, keyset-paged and capped. (permissions from EntityAnalysisModelActivationRuleOverrideService.RequiredPermissions)
- EntityAnalysisModelActivationRuleOverrideFilterFields: Lists the fields a query builder JSON filter over Activation Rule Overrides may use. (permissions from EntityAnalysisModelActivationRuleOverrideService.RequiredPermissions)
- EntityAnalysisModelActivationRuleOverrideFilter: Returns the Activation Rule Overrides matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelActivationRuleOverrideService.RequiredPermissions)
- EntityAnalysisModelActivationRuleOverrideCount: Counts the Activation Rule Overrides matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelActivationRuleOverrideService.RequiredPermissions)
- EntityAnalysisModelActivationRuleOverrideGet: Returns one Activation-Rule-scoped Override by id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelActivationRuleOverrideService.RequiredPermissions)
- EntityAnalysisModelActivationRuleOverrideGetByEntityAnalysisModelGuid: Lists Activation-Rule-scoped Overrides belonging to a given Model, scoped to the caller’s tenant. (permissions from EntityAnalysisModelActivationRuleOverrideService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Activation Rule Override Test Bench (`entity-analysis-model-activation-rule-override-test-bench`)

Hierarchy Placement: Level 4, child of Activation Supervisor
Semantic Persona: Owns the entity analysis model activation rule override slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: entity, analysis, model, activation, rule, override, test, bench
Toolbox (Max 15):
- EntityAnalysisModelActivationRuleOverrideValidate: Validates an Activation Rule Override without saving it and returns every failure. (permissions from EntityAnalysisModelActivationRuleOverrideService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Activation Rule Override Author (`entity-analysis-model-activation-rule-override-author`)

Hierarchy Placement: Level 4, child of Activation Supervisor
Semantic Persona: Owns the entity analysis model activation rule override slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: entity, analysis, model, activation, rule, override, author
Toolbox (Max 15):
- EntityAnalysisModelActivationRuleOverrideList: Lists Activation-Rule-scoped Overrides for the caller’s tenant, keyset-paged and capped. (permissions from EntityAnalysisModelActivationRuleOverrideService.RequiredPermissions)
- EntityAnalysisModelActivationRuleOverrideFilterFields: Lists the fields a query builder JSON filter over Activation Rule Overrides may use. (permissions from EntityAnalysisModelActivationRuleOverrideService.RequiredPermissions)
- EntityAnalysisModelActivationRuleOverrideFilter: Returns the Activation Rule Overrides matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelActivationRuleOverrideService.RequiredPermissions)
- EntityAnalysisModelActivationRuleOverrideCount: Counts the Activation Rule Overrides matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelActivationRuleOverrideService.RequiredPermissions)
- EntityAnalysisModelActivationRuleOverrideGet: Returns one Activation-Rule-scoped Override by id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelActivationRuleOverrideService.RequiredPermissions)
- EntityAnalysisModelActivationRuleOverrideGetByEntityAnalysisModelGuid: Lists Activation-Rule-scoped Overrides belonging to a given Model, scoped to the caller’s tenant. (permissions from EntityAnalysisModelActivationRuleOverrideService.RequiredPermissions)
- EntityAnalysisModelActivationRuleOverrideValidate: Validates an Activation Rule Override without saving it and returns every failure. (permissions from EntityAnalysisModelActivationRuleOverrideService.RequiredPermissions)
- EntityAnalysisModelActivationRuleOverrideCreate: Registers a Override against a Model/Activation Rule in the caller’s tenant; calling twice creates two. Write — confirmation required (permissions from EntityAnalysisModelActivationRuleOverrideService.RequiredPermissions)
- EntityAnalysisModelActivationRuleOverrideUpdate: Toggles a Override on or off by row existence; NOT idempotent – repeat calls flip state. Write — confirmation required (permissions from EntityAnalysisModelActivationRuleOverrideService.RequiredPermissions)
- EntityAnalysisModelActivationRuleOverrideUpdateDeleteExpiryDate: Updates the Delete Expiry Date of an existing active Override in the caller’s tenant. Write — confirmation required (permissions from EntityAnalysisModelActivationRuleOverrideService.RequiredPermissions)
- EntityAnalysisModelActivationRuleOverrideDelete: Soft-deletes a Override in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelActivationRuleOverrideService.RequiredPermissions)
- EntityAnalysisModelActivationRuleOverrideUpdateOverrideKind: Updates the Override Kind of an active Activation Rule Override in the caller’s tenant. Write — confirmation required (permissions from EntityAnalysisModelActivationRuleOverrideService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Activation Rule Override Query (`entity-analysis-model-activation-rule-override-query`)

Hierarchy Placement: Level 4, child of Activation Supervisor
Semantic Persona: Owns the entity analysis model activation rule override query slice end to end, which is small enough that reads and writes stay with one agent.
Semantic Routing Keywords: entity, analysis, model, activation, rule, override, query
Toolbox (Max 15):
- EntityAnalysisModelActivationRuleOverrideQueryGet: Lists the override-enabled Activation Rules of an Entity Analysis Model and whether each is currently overridden for a override key and value. (permissions from EntityAnalysisModelActivationRuleOverrideQueryService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Core Supervisor

Hierarchy Placement: Level 3, child of Detection Model Authoring
Semantic Persona: Routes within the core slice of detection model authoring. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: core, entity, analysis, model
Toolbox (Max 15): none — routes to 3 leaf agents covering 9 tools
Downstream Dependencies: entity-analysis-model-inspector, entity-analysis-model-test-bench, entity-analysis-model-author
Cross-Domain Reach: entity-analysis-model-synchronisation-schedule

Entity Analysis Model Inspector (`entity-analysis-model-inspector`)

Hierarchy Placement: Level 4, child of Core Supervisor
Semantic Persona: Owns the entity analysis model slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: entity, analysis, model, inspector
Toolbox (Max 15):
- EntityAnalysisModelList: Lists Models for the caller’s tenant, keyset-paged and capped. (permissions from EntityAnalysisModelService.RequiredPermissions)
- EntityAnalysisModelFilterFields: Lists the fields a query builder JSON filter over Entity Analysis Models may use. (permissions from EntityAnalysisModelService.ListPermissions)
- EntityAnalysisModelFilter: Returns the Entity Analysis Models matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelService.ListPermissions)
- EntityAnalysisModelCount: Counts the Entity Analysis Models matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelService.ListPermissions)
- EntityAnalysisModelGet: Returns one Model by id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Test Bench (`entity-analysis-model-test-bench`)

Hierarchy Placement: Level 4, child of Core Supervisor
Semantic Persona: Owns the entity analysis model slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: entity, analysis, model, test, bench
Toolbox (Max 15):
- EntityAnalysisModelValidate: Validates an Entity Analysis Model without saving it and returns every failure. ( permissions from EntityAnalysisModelService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Author (`entity-analysis-model-author`)

Hierarchy Placement: Level 4, child of Core Supervisor
Semantic Persona: Owns the entity analysis model slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: entity, analysis, model, author
Toolbox (Max 15):
- EntityAnalysisModelList: Lists Models for the caller’s tenant, keyset-paged and capped. (permissions from EntityAnalysisModelService.RequiredPermissions)
- EntityAnalysisModelFilterFields: Lists the fields a query builder JSON filter over Entity Analysis Models may use. (permissions from EntityAnalysisModelService.ListPermissions)
- EntityAnalysisModelFilter: Returns the Entity Analysis Models matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelService.ListPermissions)
- EntityAnalysisModelCount: Counts the Entity Analysis Models matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelService.ListPermissions)
- EntityAnalysisModelGet: Returns one Model by id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelService.RequiredPermissions)
- EntityAnalysisModelValidate: Validates an Entity Analysis Model without saving it and returns every failure. ( permissions from EntityAnalysisModelService.RequiredPermissions)
- EntityAnalysisModelCreate: Creates a Model in the caller’s tenant; calling twice creates two. Write — confirmation required (permissions from EntityAnalysisModelService.RequiredPermissions)
- EntityAnalysisModelUpdate: Updates a Model in the caller’s tenant by id. Write — confirmation required ( permissions from EntityAnalysisModelService.RequiredPermissions)
- EntityAnalysisModelDelete: Soft-deletes a Model in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelService.RequiredPermissions)
- EntityAnalysisModelResubmit: Resubmits a rejected Model for approval again, unchanged, opening a fresh approval cycle. Write — confirmation required (permissions from EntityAnalysisModelService.RequiredPermissions)
- EntityAnalysisModelRetract: Retracts a rejected Model, restoring it to its last approved version without a new approval cycle. Write — confirmation required (permissions from EntityAnalysisModelService.RequiredPermissions)
Downstream Dependencies: none (leaf)

References Derivation Supervisor

Hierarchy Placement: Level 3, child of Detection Model Authoring
Semantic Persona: Routes within the references derivation slice of detection model authoring. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: references derivation, entity, analysis, model, inline, function, script
Toolbox (Max 15): none — routes to 6 leaf agents covering 23 tools
Downstream Dependencies: entity-analysis-model-inline-function-inspector, entity-analysis-model-inline-function-test-bench, entity-analysis-model-inline-function-author, entity-analysis-model-inline-script-inspector, entity-analysis-model-inline-script-test-bench, entity-analysis-model-inline-script-author
Cross-Domain Reach: completions, model-invocation-context-and-dependencies

Entity Analysis Model Inline Function Inspector (`entity-analysis-model-inline-function-inspector`)

Hierarchy Placement: Level 4, child of References Derivation Supervisor
Semantic Persona: Owns the entity analysis model inline function slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: entity, analysis, model, inline, function, inspector
Toolbox (Max 15):
- EntityAnalysisModelInlineFunctionList: Lists Inline Functions for the caller’s tenant, keyset-paged and capped. (permissions from EntityAnalysisModelInlineFunctionService.ListPermissions)
- EntityAnalysisModelInlineFunctionFilterFields: Lists the fields a query builder JSON filter over Inline Functions may use. (permissions from EntityAnalysisModelInlineFunctionService.ListPermissions)
- EntityAnalysisModelInlineFunctionFilter: Returns the Inline Functions matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelInlineFunctionService.ListPermissions)
- EntityAnalysisModelInlineFunctionCount: Counts the Inline Functions matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelInlineFunctionService.ListPermissions)
- EntityAnalysisModelInlineFunctionGet: Returns one Inline Function by id, scoped to the caller’s tenant. ( permissions from EntityAnalysisModelInlineFunctionService.ReadPermissions)
- EntityAnalysisModelInlineFunctionGetByGuid: Returns one Inline Function by Guid, scoped to the caller’s tenant. (permissions from EntityAnalysisModelInlineFunctionService.ReadPermissions)
- EntityAnalysisModelInlineFunctionGetByEntityAnalysisModelId: Lists Inline Functions belonging to a given Model, scoped to the caller’s tenant. (permissions from EntityAnalysisModelInlineFunctionService.ReadPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Inline Function Test Bench (`entity-analysis-model-inline-function-test-bench`)

Hierarchy Placement: Level 4, child of References Derivation Supervisor
Semantic Persona: Owns the entity analysis model inline function slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: entity, analysis, model, inline, function, test, bench
Toolbox (Max 15):
- EntityAnalysisModelInlineFunctionValidate: Validates an Inline Function without saving it and returns every failure. (permissions from EntityAnalysisModelInlineFunctionService.WritePermissions)
- EntityAnalysisModelInlineFunctionParseRule: Parses and compiles an Inline Function’s rule text as the engine would and returns each located error. (permissions from EntityAnalysisModelInlineFunctionService.WritePermissions)
- EntityAnalysisModelInlineFunctionExecute: Runs an Inline Function against an invocation context as the engine would and returns the result. (permissions from EntityAnalysisModelInlineFunctionService.WritePermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Inline Function Author (`entity-analysis-model-inline-function-author`)

Hierarchy Placement: Level 4, child of References Derivation Supervisor
Semantic Persona: Owns the entity analysis model inline function slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: entity, analysis, model, inline, function, author
Toolbox (Max 15):
- EntityAnalysisModelInlineFunctionList: Lists Inline Functions for the caller’s tenant, keyset-paged and capped. (permissions from EntityAnalysisModelInlineFunctionService.ListPermissions)
- EntityAnalysisModelInlineFunctionFilterFields: Lists the fields a query builder JSON filter over Inline Functions may use. (permissions from EntityAnalysisModelInlineFunctionService.ListPermissions)
- EntityAnalysisModelInlineFunctionFilter: Returns the Inline Functions matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelInlineFunctionService.ListPermissions)
- EntityAnalysisModelInlineFunctionCount: Counts the Inline Functions matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelInlineFunctionService.ListPermissions)
- EntityAnalysisModelInlineFunctionGet: Returns one Inline Function by id, scoped to the caller’s tenant. ( permissions from EntityAnalysisModelInlineFunctionService.ReadPermissions)
- EntityAnalysisModelInlineFunctionGetByGuid: Returns one Inline Function by Guid, scoped to the caller’s tenant. (permissions from EntityAnalysisModelInlineFunctionService.ReadPermissions)
- EntityAnalysisModelInlineFunctionGetByEntityAnalysisModelId: Lists Inline Functions belonging to a given Model, scoped to the caller’s tenant. (permissions from EntityAnalysisModelInlineFunctionService.ReadPermissions)
- EntityAnalysisModelInlineFunctionValidate: Validates an Inline Function without saving it and returns every failure. (permissions from EntityAnalysisModelInlineFunctionService.WritePermissions)
- EntityAnalysisModelInlineFunctionParseRule: Parses and compiles an Inline Function’s rule text as the engine would and returns each located error. (permissions from EntityAnalysisModelInlineFunctionService.WritePermissions)
- EntityAnalysisModelInlineFunctionExecute: Runs an Inline Function against an invocation context as the engine would and returns the result. (permissions from EntityAnalysisModelInlineFunctionService.WritePermissions)
- EntityAnalysisModelInlineFunctionCreate: Creates an Inline Function under a Model in the caller’s tenant; calling twice creates two. Write — confirmation required (permissions from EntityAnalysisModelInlineFunctionService.WritePermissions)
- EntityAnalysisModelInlineFunctionUpdate: Updates an Inline Function in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelInlineFunctionService.WritePermissions)
- EntityAnalysisModelInlineFunctionDelete: Soft-deletes an Inline Function in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelInlineFunctionService.WritePermissions)
- EntityAnalysisModelInlineFunctionResubmit: Resubmits a rejected Inline Function for approval again, unchanged, opening a fresh approval cycle. Write — confirmation required (permissions from EntityAnalysisModelInlineFunctionService.WritePermissions)
- EntityAnalysisModelInlineFunctionRetract: Retracts a rejected Inline Function, restoring it to its last approved version without a new approval cycle. Write — confirmation required (permissions from EntityAnalysisModelInlineFunctionService.WritePermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Inline Script Inspector (`entity-analysis-model-inline-script-inspector`)

Hierarchy Placement: Level 4, child of References Derivation Supervisor
Semantic Persona: Owns the entity analysis model inline script slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: entity, analysis, model, inline, script, inspector
Toolbox (Max 15):
- EntityAnalysisModelInlineScriptList: Lists Inline Script registrations for the caller’s tenant, keyset-paged and capped. (permissions from EntityAnalysisModelInlineScriptService.ListPermissions)
- EntityAnalysisModelInlineScriptFilterFields: Lists the fields a query builder JSON filter over Inline Scripts may use. (permissions from EntityAnalysisModelInlineScriptService.ListPermissions)
- EntityAnalysisModelInlineScriptFilter: Returns the Inline Scripts matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelInlineScriptService.ListPermissions)
- EntityAnalysisModelInlineScriptCount: Counts the Inline Scripts matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelInlineScriptService.ListPermissions)
- EntityAnalysisModelInlineScriptGet: Returns one Inline Script registration by id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelInlineScriptService.ReadPermissions)
- EntityAnalysisModelInlineScriptGetByGuid: Returns one Inline Script registration by Guid, scoped to the caller’s tenant. (permissions from EntityAnalysisModelInlineScriptService.ReadPermissions)
- EntityAnalysisModelInlineScriptGetByEntityAnalysisModelId: Lists Inline Script registrations belonging to a given Model, scoped to the caller’s tenant. (permissions from EntityAnalysisModelInlineScriptService.ReadPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Inline Script Test Bench (`entity-analysis-model-inline-script-test-bench`)

Hierarchy Placement: Level 4, child of References Derivation Supervisor
Semantic Persona: Owns the entity analysis model inline script slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: entity, analysis, model, inline, script, test, bench
Toolbox (Max 15):
- EntityAnalysisModelInlineScriptValidate: Validates an Inline Script without saving it and returns every failure. (permissions from EntityAnalysisModelInlineScriptService.WritePermissions)
- EntityAnalysisModelInlineScriptExecute: Runs a model’s inline script against an invocation context when enabled; returns its properties. (permissions from EntityAnalysisModelInlineScriptService.WritePermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Inline Script Author (`entity-analysis-model-inline-script-author`)

Hierarchy Placement: Level 4, child of References Derivation Supervisor
Semantic Persona: Owns the entity analysis model inline script slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: entity, analysis, model, inline, script, author
Toolbox (Max 15):
- EntityAnalysisModelInlineScriptList: Lists Inline Script registrations for the caller’s tenant, keyset-paged and capped. (permissions from EntityAnalysisModelInlineScriptService.ListPermissions)
- EntityAnalysisModelInlineScriptFilterFields: Lists the fields a query builder JSON filter over Inline Scripts may use. (permissions from EntityAnalysisModelInlineScriptService.ListPermissions)
- EntityAnalysisModelInlineScriptFilter: Returns the Inline Scripts matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelInlineScriptService.ListPermissions)
- EntityAnalysisModelInlineScriptCount: Counts the Inline Scripts matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelInlineScriptService.ListPermissions)
- EntityAnalysisModelInlineScriptGet: Returns one Inline Script registration by id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelInlineScriptService.ReadPermissions)
- EntityAnalysisModelInlineScriptGetByGuid: Returns one Inline Script registration by Guid, scoped to the caller’s tenant. (permissions from EntityAnalysisModelInlineScriptService.ReadPermissions)
- EntityAnalysisModelInlineScriptGetByEntityAnalysisModelId: Lists Inline Script registrations belonging to a given Model, scoped to the caller’s tenant. (permissions from EntityAnalysisModelInlineScriptService.ReadPermissions)
- EntityAnalysisModelInlineScriptValidate: Validates an Inline Script without saving it and returns every failure. (permissions from EntityAnalysisModelInlineScriptService.WritePermissions)
- EntityAnalysisModelInlineScriptExecute: Runs a model’s inline script against an invocation context when enabled; returns its properties. (permissions from EntityAnalysisModelInlineScriptService.WritePermissions)
- EntityAnalysisModelInlineScriptCreate: Registers an Inline Script under a Model in the caller’s tenant; calling twice creates two. Write — confirmation required (permissions from EntityAnalysisModelInlineScriptService.WritePermissions)
- EntityAnalysisModelInlineScriptUpdate: Updates an Inline Script registration in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelInlineScriptService.WritePermissions)
- EntityAnalysisModelInlineScriptDelete: Soft-deletes an Inline Script registration in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelInlineScriptService.WritePermissions)
- EntityAnalysisModelInlineScriptResubmit: Resubmits a rejected Inline Script registration for approval again, unchanged, opening a fresh approval cycle. Write — confirmation required (permissions from EntityAnalysisModelInlineScriptService.WritePermissions)
- EntityAnalysisModelInlineScriptRetract: Retracts a rejected Inline Script registration, restoring it to its last approved version without a new approval cycle. Write — confirmation required (permissions from EntityAnalysisModelInlineScriptService.WritePermissions)
Downstream Dependencies: none (leaf)

References Gateway Supervisor

Hierarchy Placement: Level 3, child of Detection Model Authoring
Semantic Persona: Routes within the references gateway slice of detection model authoring. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: references gateway, entity, analysis, model, gateway, rule
Toolbox (Max 15): none — routes to 3 leaf agents covering 14 tools
Downstream Dependencies: entity-analysis-model-gateway-rule-inspector, entity-analysis-model-gateway-rule-test-bench, entity-analysis-model-gateway-rule-author
Cross-Domain Reach: completions, model-invocation-context-and-dependencies

Entity Analysis Model Gateway Rule Inspector (`entity-analysis-model-gateway-rule-inspector`)

Hierarchy Placement: Level 4, child of References Gateway Supervisor
Semantic Persona: Owns the entity analysis model gateway rule slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: entity, analysis, model, gateway, rule, inspector
Toolbox (Max 15):
- EntityAnalysisModelGatewayRuleList: Lists Gateway Rules for the caller’s tenant, keyset-paged and capped. ( permissions from EntityAnalysisModelGatewayRuleService.ListPermissions)
- EntityAnalysisModelGatewayRuleFilterFields: Lists the fields a query builder JSON filter over Gateway Rules may use. (permissions from EntityAnalysisModelGatewayRuleService.ListPermissions)
- EntityAnalysisModelGatewayRuleFilter: Returns the Gateway Rules matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelGatewayRuleService.ListPermissions)
- EntityAnalysisModelGatewayRuleCount: Counts the Gateway Rules matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelGatewayRuleService.ListPermissions)
- EntityAnalysisModelGatewayRuleGet: Returns one Gateway Rule by id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelGatewayRuleService.ReadPermissions)
- EntityAnalysisModelGatewayRuleGetByGuid: Returns one Gateway Rule by Guid, scoped to the caller’s tenant. ( permissions from EntityAnalysisModelGatewayRuleService.ReadPermissions)
- EntityAnalysisModelGatewayRuleGetByEntityAnalysisModelId: Lists Gateway Rules belonging to a given Model, scoped to the caller’s tenant. (permissions from EntityAnalysisModelGatewayRuleService.ReadPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Gateway Rule Test Bench (`entity-analysis-model-gateway-rule-test-bench`)

Hierarchy Placement: Level 4, child of References Gateway Supervisor
Semantic Persona: Owns the entity analysis model gateway rule slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: entity, analysis, model, gateway, rule, test, bench
Toolbox (Max 15):
- EntityAnalysisModelGatewayRuleValidate: Validates a Gateway Rule without saving it and returns every failure. ( permissions from EntityAnalysisModelGatewayRuleService.WritePermissions)
- EntityAnalysisModelGatewayRuleParseRule: Parses and compiles a Gateway Rule’s rule text as the engine would and returns each located error. (permissions from EntityAnalysisModelGatewayRuleService.WritePermissions)
- EntityAnalysisModelGatewayRuleExecute: Runs a Gateway Rule against an invocation context as the engine would and returns the result. (permissions from EntityAnalysisModelGatewayRuleService.WritePermissions)
- EntityAnalysisModelGatewayRuleBuildRuleFromBuilderJson: Turns query builder JSON into a Gateway Rule’s rule text as the browser builder would, and parses it. (permissions from EntityAnalysisModelGatewayRuleService.WritePermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Gateway Rule Author (`entity-analysis-model-gateway-rule-author`)

Hierarchy Placement: Level 4, child of References Gateway Supervisor
Semantic Persona: Owns the entity analysis model gateway rule slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: entity, analysis, model, gateway, rule, author
Toolbox (Max 15):
- EntityAnalysisModelGatewayRuleList: Lists Gateway Rules for the caller’s tenant, keyset-paged and capped. ( permissions from EntityAnalysisModelGatewayRuleService.ListPermissions)
- EntityAnalysisModelGatewayRuleFilterFields: Lists the fields a query builder JSON filter over Gateway Rules may use. (permissions from EntityAnalysisModelGatewayRuleService.ListPermissions)
- EntityAnalysisModelGatewayRuleFilter: Returns the Gateway Rules matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelGatewayRuleService.ListPermissions)
- EntityAnalysisModelGatewayRuleCount: Counts the Gateway Rules matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelGatewayRuleService.ListPermissions)
- EntityAnalysisModelGatewayRuleGet: Returns one Gateway Rule by id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelGatewayRuleService.ReadPermissions)
- EntityAnalysisModelGatewayRuleGetByGuid: Returns one Gateway Rule by Guid, scoped to the caller’s tenant. ( permissions from EntityAnalysisModelGatewayRuleService.ReadPermissions)
- EntityAnalysisModelGatewayRuleGetByEntityAnalysisModelId: Lists Gateway Rules belonging to a given Model, scoped to the caller’s tenant. (permissions from EntityAnalysisModelGatewayRuleService.ReadPermissions)
- EntityAnalysisModelGatewayRuleValidate: Validates a Gateway Rule without saving it and returns every failure. ( permissions from EntityAnalysisModelGatewayRuleService.WritePermissions)
- EntityAnalysisModelGatewayRuleParseRule: Parses and compiles a Gateway Rule’s rule text as the engine would and returns each located error. (permissions from EntityAnalysisModelGatewayRuleService.WritePermissions)
- EntityAnalysisModelGatewayRuleExecute: Runs a Gateway Rule against an invocation context as the engine would and returns the result. (permissions from EntityAnalysisModelGatewayRuleService.WritePermissions)
- EntityAnalysisModelGatewayRuleBuildRuleFromBuilderJson: Turns query builder JSON into a Gateway Rule’s rule text as the browser builder would, and parses it. (permissions from EntityAnalysisModelGatewayRuleService.WritePermissions)
- EntityAnalysisModelGatewayRuleCreate: Registers a Gateway Rule under a Model in the caller’s tenant; calling twice creates two. Write — confirmation required (permissions from EntityAnalysisModelGatewayRuleService.WritePermissions)
- EntityAnalysisModelGatewayRuleUpdate: Updates a Gateway Rule in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelGatewayRuleService.WritePermissions)
- EntityAnalysisModelGatewayRuleDelete: Soft-deletes a Gateway Rule in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelGatewayRuleService.WritePermissions)
- EntityAnalysisModelGatewayRuleResetCounter: Resets the activation and evaluation counters for a Gateway Rule in the caller’s tenant to zero. Write — confirmation required (permissions from EntityAnalysisModelGatewayRuleService.WritePermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Gateway Rule Approvals (`entity-analysis-model-gateway-rule-approvals`)

Hierarchy Placement: Level 4, child of References Gateway Supervisor
Semantic Persona: Owns the maker-checker aftermath of the entity analysis model gateway rule slice: resolves a rejected submission by resubmitting it unchanged or retracting it to its last approved version. Never creates, updates or deletes it.
Semantic Routing Keywords: entity, analysis, model, gateway, rule, approvals, resubmit, retract
Toolbox (Max 15):
- EntityAnalysisModelGatewayRuleResubmit: Resubmits a rejected Gateway Rule for approval again, unchanged, opening a fresh approval cycle. Write — confirmation required (permissions from EntityAnalysisModelGatewayRuleService.WritePermissions)
- EntityAnalysisModelGatewayRuleRetract: Retracts a rejected Gateway Rule, restoring it to its last approved version without a new approval cycle. Write — confirmation required (permissions from EntityAnalysisModelGatewayRuleService.WritePermissions)
Downstream Dependencies: none (leaf)

References Payload Contract Supervisor

Hierarchy Placement: Level 3, child of Detection Model Authoring
Semantic Persona: Routes within the references payload contract slice of detection model authoring. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: references payload contract, entity, analysis, model, request, path, inline, script, potential, multi, part, string, names, by, integer, float, data, type
Toolbox (Max 15): none — routes to 5 leaf agents covering 20 tools
Downstream Dependencies: entity-analysis-model-request-x-path-inspector-1, entity-analysis-model-request-x-path-inspector-2, entity-analysis-model-request-x-path-test-bench, entity-analysis-model-request-x-path-author, entity-analysis-observability

Entity Analysis Model Request X Path Inspector 1 (`entity-analysis-model-request-x-path-inspector-1`)

Hierarchy Placement: Level 4, child of References Payload Contract Supervisor
Semantic Persona: Owns the entity analysis model request x path slice end to end, which is small enough that reads and writes stay with one agent.
Semantic Routing Keywords: entity, analysis, model, request, path, inspector, 1
Toolbox (Max 15):
- EntityAnalysisModelRequestXPathList: Lists Request XPaths for the caller’s tenant, keyset-paged and capped. ( permissions from EntityAnalysisModelRequestXPathService.ListPermissions)
- EntityAnalysisModelRequestXPathFilterFields: Lists the fields a query builder JSON filter over RequestXPaths may use. (permissions from EntityAnalysisModelRequestXPathService.ListPermissions)
- EntityAnalysisModelRequestXPathFilter: Returns the RequestXPaths matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelRequestXPathService.ListPermissions)
- EntityAnalysisModelRequestXPathCount: Counts the RequestXPaths matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelRequestXPathService.ListPermissions)
- EntityAnalysisModelRequestXPathGet: Returns one Request XPath by id, scoped to the caller’s tenant. ( permissions from EntityAnalysisModelRequestXPathService.ReadPermissions)
- EntityAnalysisModelRequestXPathGetByEntityAnalysisModelId: Lists Request XPaths belonging to a given Model, scoped to the caller’s tenant. (permissions from EntityAnalysisModelRequestXPathService.ReadByEntityAnalysisModelPermissions)
- EntityAnalysisModelRequestXPathGetByCasesWorkflowId: Lists Request XPaths reachable from a given Case Workflow’s Model, scoped to the caller’s tenant. (permissions from EntityAnalysisModelRequestXPathService.ReadPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Request X Path Inspector 2 (`entity-analysis-model-request-x-path-inspector-2`)

Hierarchy Placement: Level 4, child of References Payload Contract Supervisor
Semantic Persona: Owns the entity analysis model request x path slice end to end, which is small enough that reads and writes stay with one agent.
Semantic Routing Keywords: entity, analysis, model, request, path, inspector, 2
Toolbox (Max 15):
- EntityAnalysisModelRequestXPathGetByOverrideKey: Lists Request XPaths flagged for Override, scoped to the caller’s tenant. (permissions from EntityAnalysisModelRequestXPathService.OverridePermissions)
- EntityAnalysisModelRequestXPathGetByEntityAnalysisModelIdByDataType: Lists String, Integer or Float Request XPaths belonging to a given Model, scoped to the caller’s tenant. (permissions from EntityAnalysisModelRequestXPathService.ReadByDataTypePermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Request X Path Test Bench (`entity-analysis-model-request-x-path-test-bench`)

Hierarchy Placement: Level 4, child of References Payload Contract Supervisor
Semantic Persona: Owns the entity analysis model request x path slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: entity, analysis, model, request, path, test, bench
Toolbox (Max 15):
- EntityAnalysisModelRequestXPathValidate: Validates a RequestXPath without saving it and returns every failure. (permissions from EntityAnalysisModelRequestXPathService.WritePermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Request X Path Author (`entity-analysis-model-request-x-path-author`)

Hierarchy Placement: Level 4, child of References Payload Contract Supervisor
Semantic Persona: Owns the entity analysis model request x path slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: entity, analysis, model, request, path, author
Toolbox (Max 15):
- EntityAnalysisModelRequestXPathList: Lists Request XPaths for the caller’s tenant, keyset-paged and capped. ( permissions from EntityAnalysisModelRequestXPathService.ListPermissions)
- EntityAnalysisModelRequestXPathFilterFields: Lists the fields a query builder JSON filter over RequestXPaths may use. (permissions from EntityAnalysisModelRequestXPathService.ListPermissions)
- EntityAnalysisModelRequestXPathFilter: Returns the RequestXPaths matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelRequestXPathService.ListPermissions)
- EntityAnalysisModelRequestXPathCount: Counts the RequestXPaths matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelRequestXPathService.ListPermissions)
- EntityAnalysisModelRequestXPathGet: Returns one Request XPath by id, scoped to the caller’s tenant. ( permissions from EntityAnalysisModelRequestXPathService.ReadPermissions)
- EntityAnalysisModelRequestXPathGetByEntityAnalysisModelId: Lists Request XPaths belonging to a given Model, scoped to the caller’s tenant. (permissions from EntityAnalysisModelRequestXPathService.ReadByEntityAnalysisModelPermissions)
- EntityAnalysisModelRequestXPathGetByCasesWorkflowId: Lists Request XPaths reachable from a given Case Workflow’s Model, scoped to the caller’s tenant. (permissions from EntityAnalysisModelRequestXPathService.ReadPermissions)
- EntityAnalysisModelRequestXPathGetByOverrideKey: Lists Request XPaths flagged for Override, scoped to the caller’s tenant. (permissions from EntityAnalysisModelRequestXPathService.OverridePermissions)
- EntityAnalysisModelRequestXPathGetByEntityAnalysisModelIdByDataType: Lists String, Integer or Float Request XPaths belonging to a given Model, scoped to the caller’s tenant. (permissions from EntityAnalysisModelRequestXPathService.ReadByDataTypePermissions)
- EntityAnalysisModelRequestXPathValidate: Validates a RequestXPath without saving it and returns every failure. (permissions from EntityAnalysisModelRequestXPathService.WritePermissions)
- EntityAnalysisModelRequestXPathCreate: Creates a Request XPath under a Model in the caller’s tenant; calling twice creates two. Write — confirmation required (permissions from EntityAnalysisModelRequestXPathService.WritePermissions)
- EntityAnalysisModelRequestXPathUpdate: Updates a Request XPath in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelRequestXPathService.WritePermissions)
- EntityAnalysisModelRequestXPathDelete: Soft-deletes a Request XPath in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelRequestXPathService.WritePermissions)
- EntityAnalysisModelRequestXPathResubmit: Resubmits a rejected Request XPath for approval again, unchanged, opening a fresh approval cycle. Write — confirmation required (permissions from EntityAnalysisModelRequestXPathService.WritePermissions)
- EntityAnalysisModelRequestXPathRetract: Retracts a rejected Request XPath, restoring it to its last approved version without a new approval cycle. Write — confirmation required (permissions from EntityAnalysisModelRequestXPathService.WritePermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Observability (`entity-analysis-observability`)

Hierarchy Placement: Level 4, child of References Payload Contract Supervisor
Semantic Persona: Owns the entity analysis observability slice end to end, which is small enough that reads and writes stay with one agent.
Semantic Routing Keywords: entity, analysis, inline, script, potential, multi, part, string, names, request, path, by, integer, float, data, type
Toolbox (Max 15):
- EntityAnalysisInlineScriptList: Lists registered Inline Scripts from the system-wide catalogue, keyset-paged and capped. (permissions from EntityAnalysisInlineScriptService.ListPermissions)
- EntityAnalysisInlineScriptFilterFields: Lists the fields a query builder JSON filter over Entity Analysis Inline Scripts may use. (permissions from EntityAnalysisInlineScriptService.ListPermissions)
- EntityAnalysisInlineScriptFilter: Returns the Entity Analysis Inline Scripts matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisInlineScriptService.ListPermissions)
- EntityAnalysisInlineScriptCount: Counts the Entity Analysis Inline Scripts matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisInlineScriptService.ListPermissions)
- EntityAnalysisPotentialMultiPartStringNamesGetById: Lists the names of potential multi-part string fields for an entity analysis model, by model Id. (permissions from EntityAnalysisPotentialMultiPartStringNamesService.RequiredPermissions)
- EntityAnalysisPotentialMultiPartStringNamesGetByGuid: Lists the names of potential multi-part string fields for an entity analysis model, by model Guid. (permissions from EntityAnalysisPotentialMultiPartStringNamesService.RequiredPermissions)
- EntityAnalysisRequestXPathInlineScriptNamesByStringIntegerFloatDataTypeGet: Lists the String and Float fields of an Entity Analysis Model usable as search keys. (permissions from EntityAnalysisRequestXPathInlineScriptNamesByStringIntegerFloatDataTypeService.RequiredPermissions)
Downstream Dependencies: none (leaf)

References Watchlists Supervisor

Hierarchy Placement: Level 3, child of Detection Model Authoring
Semantic Persona: Routes within the references watchlists slice of detection model authoring. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: references watchlists, entity, analysis, model, sanction, tag
Toolbox (Max 15): none — routes to 6 leaf agents covering 20 tools
Downstream Dependencies: entity-analysis-model-sanction-inspector, entity-analysis-model-sanction-test-bench, entity-analysis-model-sanction-author, entity-analysis-model-tag-inspector, entity-analysis-model-tag-test-bench, entity-analysis-model-tag-author
Cross-Domain Reach: completions

Entity Analysis Model Sanction Inspector (`entity-analysis-model-sanction-inspector`)

Hierarchy Placement: Level 4, child of References Watchlists Supervisor
Semantic Persona: Owns the entity analysis model sanction slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: entity, analysis, model, sanction, inspector
Toolbox (Max 15):
- EntityAnalysisModelSanctionList: Lists Sanction checks for the caller’s tenant, keyset-paged and capped. ( permissions from EntityAnalysisModelSanctionService.ListPermissions)
- EntityAnalysisModelSanctionFilterFields: Lists the fields a query builder JSON filter over Sanctions may use. ( permissions from EntityAnalysisModelSanctionService.ListPermissions)
- EntityAnalysisModelSanctionFilter: Returns the Sanctions matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelSanctionService.ListPermissions)
- EntityAnalysisModelSanctionCount: Counts the Sanctions matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelSanctionService.ListPermissions)
- EntityAnalysisModelSanctionGet: Returns one Sanction check by id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelSanctionService.ReadPermissions)
- EntityAnalysisModelSanctionGetByGuid: Returns one Sanction check by Guid, scoped to the caller’s tenant. ( permissions from EntityAnalysisModelSanctionService.ReadPermissions)
- EntityAnalysisModelSanctionGetByEntityAnalysisModelId: Lists Sanction checks belonging to a given Model, scoped to the caller’s tenant. (permissions from EntityAnalysisModelSanctionService.ReadPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Sanction Test Bench (`entity-analysis-model-sanction-test-bench`)

Hierarchy Placement: Level 4, child of References Watchlists Supervisor
Semantic Persona: Owns the entity analysis model sanction slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: entity, analysis, model, sanction, test, bench
Toolbox (Max 15):
- EntityAnalysisModelSanctionValidate: Validates a Sanction without saving it and returns every failure. ( permissions from EntityAnalysisModelSanctionService.WritePermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Sanction Author (`entity-analysis-model-sanction-author`)

Hierarchy Placement: Level 4, child of References Watchlists Supervisor
Semantic Persona: Owns the entity analysis model sanction slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: entity, analysis, model, sanction, author
Toolbox (Max 15):
- EntityAnalysisModelSanctionList: Lists Sanction checks for the caller’s tenant, keyset-paged and capped. ( permissions from EntityAnalysisModelSanctionService.ListPermissions)
- EntityAnalysisModelSanctionFilterFields: Lists the fields a query builder JSON filter over Sanctions may use. ( permissions from EntityAnalysisModelSanctionService.ListPermissions)
- EntityAnalysisModelSanctionFilter: Returns the Sanctions matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelSanctionService.ListPermissions)
- EntityAnalysisModelSanctionCount: Counts the Sanctions matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelSanctionService.ListPermissions)
- EntityAnalysisModelSanctionGet: Returns one Sanction check by id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelSanctionService.ReadPermissions)
- EntityAnalysisModelSanctionGetByGuid: Returns one Sanction check by Guid, scoped to the caller’s tenant. ( permissions from EntityAnalysisModelSanctionService.ReadPermissions)
- EntityAnalysisModelSanctionGetByEntityAnalysisModelId: Lists Sanction checks belonging to a given Model, scoped to the caller’s tenant. (permissions from EntityAnalysisModelSanctionService.ReadPermissions)
- EntityAnalysisModelSanctionValidate: Validates a Sanction without saving it and returns every failure. ( permissions from EntityAnalysisModelSanctionService.WritePermissions)
- EntityAnalysisModelSanctionCreate: Registers a Sanction check under a Model in the caller’s tenant; calling twice creates two. Write — confirmation required (permissions from EntityAnalysisModelSanctionService.WritePermissions)
- EntityAnalysisModelSanctionUpdate: Updates a Sanction check in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelSanctionService.WritePermissions)
- EntityAnalysisModelSanctionDelete: Soft-deletes a Sanction check in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelSanctionService.WritePermissions)
- EntityAnalysisModelSanctionResubmit: Resubmits a rejected Sanction check for approval again, unchanged, opening a fresh approval cycle. Write — confirmation required (permissions from EntityAnalysisModelSanctionService.WritePermissions)
- EntityAnalysisModelSanctionRetract: Retracts a rejected Sanction check, restoring it to its last approved version without a new approval cycle. Write — confirmation required (permissions from EntityAnalysisModelSanctionService.WritePermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Tag Inspector (`entity-analysis-model-tag-inspector`)

Hierarchy Placement: Level 4, child of References Watchlists Supervisor
Semantic Persona: Owns the entity analysis model tag slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: entity, analysis, model, tag, inspector
Toolbox (Max 15):
- EntityAnalysisModelTagList: Lists Tags for the caller’s tenant, keyset-paged and capped. (permissions from EntityAnalysisModelTagService.RequiredPermissions)
- EntityAnalysisModelTagFilterFields: Lists the fields a query builder JSON filter over Tags may use. ( permissions from EntityAnalysisModelTagService.RequiredPermissions)
- EntityAnalysisModelTagFilter: Returns the Tags matching query builder JSON, keyset-paged and capped. ( permissions from EntityAnalysisModelTagService.RequiredPermissions)
- EntityAnalysisModelTagCount: Counts the Tags matching query builder JSON, optionally grouped by a field. ( permissions from EntityAnalysisModelTagService.RequiredPermissions)
- EntityAnalysisModelTagGet: Returns one Tag by id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelTagService.RequiredPermissions)
- EntityAnalysisModelTagGetByGuid: Returns one Tag by Guid, scoped to the caller’s tenant. (permissions from EntityAnalysisModelTagService.RequiredPermissions)
- EntityAnalysisModelTagGetByEntityAnalysisModelId: Lists the Active Tags belonging to a given Model, ordered by Name, scoped to the caller’s tenant. (permissions from EntityAnalysisModelTagService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Tag Test Bench (`entity-analysis-model-tag-test-bench`)

Hierarchy Placement: Level 4, child of References Watchlists Supervisor
Semantic Persona: Owns the entity analysis model tag slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: entity, analysis, model, tag, test, bench
Toolbox (Max 15):
- EntityAnalysisModelTagValidate: Validates a Tag without saving it and returns every failure. (permissions from EntityAnalysisModelTagService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Tag Author (`entity-analysis-model-tag-author`)

Hierarchy Placement: Level 4, child of References Watchlists Supervisor
Semantic Persona: Owns the entity analysis model tag slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: entity, analysis, model, tag, author
Toolbox (Max 15):
- EntityAnalysisModelTagList: Lists Tags for the caller’s tenant, keyset-paged and capped. (permissions from EntityAnalysisModelTagService.RequiredPermissions)
- EntityAnalysisModelTagFilterFields: Lists the fields a query builder JSON filter over Tags may use. ( permissions from EntityAnalysisModelTagService.RequiredPermissions)
- EntityAnalysisModelTagFilter: Returns the Tags matching query builder JSON, keyset-paged and capped. ( permissions from EntityAnalysisModelTagService.RequiredPermissions)
- EntityAnalysisModelTagCount: Counts the Tags matching query builder JSON, optionally grouped by a field. ( permissions from EntityAnalysisModelTagService.RequiredPermissions)
- EntityAnalysisModelTagGet: Returns one Tag by id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelTagService.RequiredPermissions)
- EntityAnalysisModelTagGetByGuid: Returns one Tag by Guid, scoped to the caller’s tenant. (permissions from EntityAnalysisModelTagService.RequiredPermissions)
- EntityAnalysisModelTagGetByEntityAnalysisModelId: Lists the Active Tags belonging to a given Model, ordered by Name, scoped to the caller’s tenant. (permissions from EntityAnalysisModelTagService.RequiredPermissions)
- EntityAnalysisModelTagValidate: Validates a Tag without saving it and returns every failure. (permissions from EntityAnalysisModelTagService.RequiredPermissions)
- EntityAnalysisModelTagCreate: Registers a Tag under a Model in the caller’s tenant; calling twice creates two. Write — confirmation required (permissions from EntityAnalysisModelTagService.RequiredPermissions)
- EntityAnalysisModelTagUpdate: Updates a Tag in the caller’s tenant by id. Write — confirmation required ( permissions from EntityAnalysisModelTagService.RequiredPermissions)
- EntityAnalysisModelTagDelete: Soft-deletes a Tag in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelTagService.RequiredPermissions)
- EntityAnalysisModelTagResubmit: Resubmits a rejected Tag for approval again, unchanged, opening a fresh approval cycle. Write — confirmation required (permissions from EntityAnalysisModelTagService.RequiredPermissions)
- EntityAnalysisModelTagRetract: Retracts a rejected Tag, restoring it to its last approved version without a new approval cycle. Write — confirmation required (permissions from EntityAnalysisModelTagService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Model Risk and Analytics (`model-risk`)

Model Risk and Analytics

Hierarchy Placement: Level 2, child of Jooby
Semantic Persona: Owns everything that trains, validates, promotes and explains a model: Exhaustive search, Advanced Analytics, backtesting, integrity and HTTP adaptation. This is the model-risk-management function and the only place a model’s predictive behaviour changes.
Semantic Routing Keywords: train, training run, promote, rollback, backtest, calibration, anomaly, classifier, regressor, clusterer, model integrity, drift, explainability
Toolbox (Max 15): none — routes to 4 supervisors and 1 leaf agent covering 133 tools
Downstream Dependencies: model-risk.advanced-analytics, model-risk.advanced-analytics-training, model-risk.exhaustive, model-risk.validation, advanced-analytics-training-data
Cross-Domain Reach: completions

Advanced Analytics Training Data (`advanced-analytics-training-data`)

Hierarchy Placement: Level 3, child of Model Risk and Analytics
Semantic Persona: Turns a table somebody has queried into the numeric training data a run reads, choosing columns by name and packing the outcome last. It trains nothing. It hangs off the domain rather than off the training supervisor because a leaf cannot delegate, so a pointer prepared below the supervisor could never be handed back up to it.
Semantic Routing Keywords: training data, data pointer, cache pointer, label column, predictors, prepare, matrix
Toolbox (Max 15):
- AdvancedAnalyticsTrainingDataPut: Turns a cached table into the numeric training data a run reads, by named columns, and returns the cache pointer to submit against. Write — confirmation required (permissions from AdvancedAnalyticsPermissionSpecifications.TrainingOnly)
- AdvancedAnalyticsTrainingDataGet: Reports the shape, size and expiry of training data already put in the cache, or nothing when it has expired. (permissions from AdvancedAnalyticsPermissionSpecifications.AnalyticsOnly)
Downstream Dependencies: none (leaf)

Advanced Analytics Supervisor

Hierarchy Placement: Level 3, child of Model Risk and Analytics
Semantic Persona: Routes within the advanced analytics slice of model risk and analytics. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: advanced analytics, advanced, analytics, instance, model
Toolbox (Max 15): none — routes to 5 leaf agents covering 19 tools
Downstream Dependencies: advanced-analytics-instance-inspector, advanced-analytics-instance-test-bench, advanced-analytics-instance-author, advanced-analytics-model-inspector, advanced-analytics-model-author

Advanced Analytics Instance Inspector (`advanced-analytics-instance-inspector`)

Hierarchy Placement: Level 4, child of Advanced Analytics Supervisor
Semantic Persona: Owns the advanced analytics instance slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: advanced, analytics, instance, inspector
Toolbox (Max 15):
- AdvancedAnalyticsInstanceGet: Returns one Advanced Analytics instance by id, scoped to the caller’s tenant. ( permissions from AdvancedAnalyticsPermissionSpecifications.AnalyticsOnly)
- AdvancedAnalyticsInstanceGetByGuid: Returns one Advanced Analytics instance by Guid, scoped to the caller’s tenant. (permissions from AdvancedAnalyticsPermissionSpecifications.AnalyticsOnly)
- AdvancedAnalyticsInstanceList: Lists Advanced Analytics instances for the caller’s tenant, keyset-paged and capped. (permissions from AdvancedAnalyticsPermissionSpecifications.AnalyticsOnly)
- AdvancedAnalyticsInstanceFilterFields: Lists the fields a query builder JSON filter over Advanced Analytics instances may use. (permissions from AdvancedAnalyticsPermissionSpecifications.AnalyticsOnly)
- AdvancedAnalyticsInstanceFilter: Returns the Advanced Analytics instances matching query builder JSON, keyset-paged and capped. (permissions from AdvancedAnalyticsPermissionSpecifications.AnalyticsOnly)
- AdvancedAnalyticsInstanceCount: Counts the Advanced Analytics instances matching query builder JSON, optionally grouped by a field. (permissions from AdvancedAnalyticsPermissionSpecifications.AnalyticsOnly)
Downstream Dependencies: none (leaf)

Advanced Analytics Instance Test Bench (`advanced-analytics-instance-test-bench`)

Hierarchy Placement: Level 4, child of Advanced Analytics Supervisor
Semantic Persona: Owns the advanced analytics instance slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: advanced, analytics, instance, test, bench
Toolbox (Max 15):
- AdvancedAnalyticsInstanceValidate: Validates an Advanced Analytics instance without saving it and returns every failure. (permissions from AdvancedAnalyticsPermissionSpecifications.AnalyticsOnly)
Downstream Dependencies: none (leaf)

Advanced Analytics Instance Author (`advanced-analytics-instance-author`)

Hierarchy Placement: Level 4, child of Advanced Analytics Supervisor
Semantic Persona: Owns the advanced analytics instance slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: advanced, analytics, instance, author
Toolbox (Max 15):
- AdvancedAnalyticsInstanceGet: Returns one Advanced Analytics instance by id, scoped to the caller’s tenant. ( permissions from AdvancedAnalyticsPermissionSpecifications.AnalyticsOnly)
- AdvancedAnalyticsInstanceGetByGuid: Returns one Advanced Analytics instance by Guid, scoped to the caller’s tenant. (permissions from AdvancedAnalyticsPermissionSpecifications.AnalyticsOnly)
- AdvancedAnalyticsInstanceList: Lists Advanced Analytics instances for the caller’s tenant, keyset-paged and capped. (permissions from AdvancedAnalyticsPermissionSpecifications.AnalyticsOnly)
- AdvancedAnalyticsInstanceFilterFields: Lists the fields a query builder JSON filter over Advanced Analytics instances may use. (permissions from AdvancedAnalyticsPermissionSpecifications.AnalyticsOnly)
- AdvancedAnalyticsInstanceFilter: Returns the Advanced Analytics instances matching query builder JSON, keyset-paged and capped. (permissions from AdvancedAnalyticsPermissionSpecifications.AnalyticsOnly)
- AdvancedAnalyticsInstanceCount: Counts the Advanced Analytics instances matching query builder JSON, optionally grouped by a field. (permissions from AdvancedAnalyticsPermissionSpecifications.AnalyticsOnly)
- AdvancedAnalyticsInstanceValidate: Validates an Advanced Analytics instance without saving it and returns every failure. (permissions from AdvancedAnalyticsPermissionSpecifications.AnalyticsOnly)
- AdvancedAnalyticsInstanceCreate: Creates an unallocated Advanced Analytics instance in the caller’s tenant. Write — confirmation required (permissions from AdvancedAnalyticsPermissionSpecifications.AnalyticsOnly)
- AdvancedAnalyticsInstanceUpdate: Renames an instance or changes its output flags; never changes its allocation. Write — confirmation required (permissions from AdvancedAnalyticsPermissionSpecifications.AnalyticsOnly)
- AdvancedAnalyticsInstanceDelete: Soft-deletes an Advanced Analytics instance, taking it out of the live pipeline. Write — confirmation required (permissions from AdvancedAnalyticsPermissionSpecifications.AnalyticsOnly)
- AdvancedAnalyticsInstanceAllocateToModel: Allocates an instance to a Jube Model of the caller’s tenant; needs the Promote Model permission. Write — confirmation required (permissions from AdvancedAnalyticsPermissionSpecifications.PromoteModelOnly)
- AdvancedAnalyticsInstanceDeallocate: Removes an instance’s allocation to a Jube Model; needs the Promote Model permission. Write — confirmation required (permissions from AdvancedAnalyticsPermissionSpecifications.PromoteModelOnly)
- AdvancedAnalyticsInstanceResubmit: Resubmits a rejected change to an instance for approval again; needs the Promote Model permission. Write — confirmation required (permissions from AdvancedAnalyticsPermissionSpecifications.PromoteModelOnly)
- AdvancedAnalyticsInstanceRetract: Retracts a rejected change to an instance, restoring its last approved state; needs the Promote Model permission. Write — confirmation required (permissions from AdvancedAnalyticsPermissionSpecifications.PromoteModelOnly)
Downstream Dependencies: none (leaf)

Advanced Analytics Model Inspector (`advanced-analytics-model-inspector`)

Hierarchy Placement: Level 4, child of Advanced Analytics Supervisor
Semantic Persona: Owns the advanced analytics model slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: advanced, analytics, model, inspector
Toolbox (Max 15):
- AdvancedAnalyticsModelList: Lists trained models for the caller’s tenant with their live version, keyset-paged and capped. (permissions from AdvancedAnalyticsPermissionSpecifications.AnalyticsOnly)
- AdvancedAnalyticsModelGet: Returns one trained model with all its versions and their headline metrics. ( permissions from AdvancedAnalyticsPermissionSpecifications.AnalyticsOnly)
- AdvancedAnalyticsModelListVersions: Lists the versions of a trained model with headline metrics and warning codes. (permissions from AdvancedAnalyticsPermissionSpecifications.AnalyticsOnly)
- AdvancedAnalyticsModelGetVersion: Returns one model version with its complete stored evidence. (permissions from AdvancedAnalyticsPermissionSpecifications.AnalyticsOnly)
- AdvancedAnalyticsModelCompareVersions: Compares two model versions on held-out metrics and says whether the comparison is fair. (permissions from AdvancedAnalyticsPermissionSpecifications.AnalyticsOnly)
- AdvancedAnalyticsModelGetPromotionEvidence: Returns the evidence needed to decide whether a model version deserves promotion. (permissions from AdvancedAnalyticsPermissionSpecifications.AnalyticsOnly)
Downstream Dependencies: none (leaf)

Advanced Analytics Model Author (`advanced-analytics-model-author`)

Hierarchy Placement: Level 4, child of Advanced Analytics Supervisor
Semantic Persona: Owns the advanced analytics model slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: advanced, analytics, model, author
Toolbox (Max 15):
- AdvancedAnalyticsModelList: Lists trained models for the caller’s tenant with their live version, keyset-paged and capped. (permissions from AdvancedAnalyticsPermissionSpecifications.AnalyticsOnly)
- AdvancedAnalyticsModelGet: Returns one trained model with all its versions and their headline metrics. ( permissions from AdvancedAnalyticsPermissionSpecifications.AnalyticsOnly)
- AdvancedAnalyticsModelListVersions: Lists the versions of a trained model with headline metrics and warning codes. (permissions from AdvancedAnalyticsPermissionSpecifications.AnalyticsOnly)
- AdvancedAnalyticsModelGetVersion: Returns one model version with its complete stored evidence. (permissions from AdvancedAnalyticsPermissionSpecifications.AnalyticsOnly)
- AdvancedAnalyticsModelCompareVersions: Compares two model versions on held-out metrics and says whether the comparison is fair. (permissions from AdvancedAnalyticsPermissionSpecifications.AnalyticsOnly)
- AdvancedAnalyticsModelGetPromotionEvidence: Returns the evidence needed to decide whether a model version deserves promotion. (permissions from AdvancedAnalyticsPermissionSpecifications.AnalyticsOnly)
- AdvancedAnalyticsModelPromoteVersion: Makes a model version live, recording its evidence; needs the Promote Model permission. Write — confirmation required (permissions from AdvancedAnalyticsPermissionSpecifications.PromoteModelOnly)
- AdvancedAnalyticsModelRollback: Puts a model back on an earlier version; needs the Promote Model permission. Write — confirmation required (permissions from AdvancedAnalyticsPermissionSpecifications.PromoteModelOnly)
Downstream Dependencies: none (leaf)

Advanced Analytics Training Supervisor

Hierarchy Placement: Level 3, child of Model Risk and Analytics
Semantic Persona: Routes within the advanced analytics training slice of model risk and analytics by academic model kind rather than by what sounds closest to the request. Holds no tools; every run below belongs to an Advanced Analytics instance that must already exist, and training data must already be a cache pointer (via the training data specialist) before a run can be opened.
Semantic Routing Keywords: train, training run, classifier, regressor, clusterer, anomaly, latent structure, model family, academic kind
Toolbox (Max 15): none — routes to 8 leaf agents covering 56 tools
Downstream Dependencies: advanced-analytics-training-classifier-trees, advanced-analytics-training-classifier-probabilistic, advanced-analytics-training-classifier-margin, advanced-analytics-training-clusterer-centroid, advanced-analytics-training-clusterer-density, advanced-analytics-training-regressor, advanced-analytics-training-anomaly, advanced-analytics-training-latent-structure
Cross-Domain Reach: advanced-analytics-instance-inspector

Advanced Analytics Training Classifier Trees (`advanced-analytics-training-classifier-trees`)

Hierarchy Placement: Level 4, child of Advanced Analytics Training Supervisor
Semantic Persona: Trains the tree and ensemble classifiers of advanced analytics: a single decision tree, a random forest, a bagged ensemble, AdaBoost and gradient boosting. All five fit axis-parallel splits and differ in how many trees they grow and how they weight them.
Semantic Routing Keywords: decision tree, random forest, bagging, adaboost, gradient boosting, classifier, ensemble, tree
Toolbox (Max 15):
- SubmitDecisionTreeClassifierTraining: Queues a decision tree classification training run: fits nested yes/no threshold splits that best separate the event of interest, with the leaf’s calibrated share of positive rows as the recalled probability. (permissions from AdvancedAnalyticsPermissions.Training)
- SubmitRandomForestClassifierTraining: Queues a random forest classification training run: averages many bootstrap-sampled decision trees’ votes into a steadier, Platt-calibrated probability than any single tree’s. ( permissions from AdvancedAnalyticsPermissions.Training)
- SubmitBaggingClassifierTraining: Queues a bagging classification training run: averages many unrestricted decision trees, each grown on its own bootstrap sample, into a Platt-calibrated probability. (permissions from AdvancedAnalyticsPermissions.Training)
- SubmitAdaBoostClassifierTraining: Queues an AdaBoost classification training run: fits shallow trees one after another, each reweighted onto the previous rounds’ mistakes, and Platt-calibrates the alpha-weighted vote margin. (permissions from AdvancedAnalyticsPermissions.Training)
- SubmitGradientBoostingClassifierTraining: Queues a gradient boosting classification training run: fits shallow trees successively to the current log-odds residual and Platt-calibrates the resulting margin. (permissions from AdvancedAnalyticsPermissions.Training)
- GetAdvancedAnalyticsTrainingRunStatus: Reads where a queued training run has got to: Pending, Running, then Succeeded, Failed or Cancelled; works for a run submitted by any family’s Submit tool. (permissions from AdvancedAnalyticsPermissions.Analytics)
- CancelAdvancedAnalyticsTrainingRun: Cancels a training run submitted by any family’s Submit tool; a finished run is returned unchanged. (permissions from AdvancedAnalyticsPermissions.Training)
- GetAdvancedAnalyticsTrainingRunResult: Reads the outcome of a finished training run: its final status and, once succeeded, the trained model and version plus family-specific evidence. (permissions from AdvancedAnalyticsPermissions.Analytics)
Downstream Dependencies: none (leaf)

Advanced Analytics Training Classifier Probabilistic (`advanced-analytics-training-classifier-probabilistic`)

Hierarchy Placement: Level 4, child of Advanced Analytics Training Supervisor
Semantic Persona: Trains the probabilistic classifiers of advanced analytics: logistic regression, naive Bayes and tree-augmented naive Bayes. These three estimate a probability as part of the fit rather than having one scaled on afterwards, so their evidence reads as coefficients or conditional tables rather than as a vote.
Semantic Routing Keywords: logistic regression, naive bayes, tree augmented naive bayes, probabilistic classifier, coefficients
Toolbox (Max 15):
- SubmitLogisticRegressionTraining: Queues a logistic regression training run against inline rows or a cache pointer, fitting a ridge-penalised coefficient per predictor. (permissions from AdvancedAnalyticsPermissions.Training)
- SubmitNaiveBayesClassifierTraining: Queues a Gaussian naive Bayes classifier training run, fitting one distribution per predictor per class and Platt-calibrating the resulting posterior. (permissions from AdvancedAnalyticsPermissions.Training)
- SubmitTreeAugmentedNaiveBayesClassifierTraining: Queues a tree-augmented naive Bayes training run: a Chow-Liu-discovered one-parent-per-predictor structure relaxes naive Bayes’ independence assumption, with the posterior Platt-calibrated. (permissions from AdvancedAnalyticsPermissions.Training)
- GetAdvancedAnalyticsTrainingRunStatus: Reads where a queued training run has got to: Pending, Running, then Succeeded, Failed or Cancelled; works for a run submitted by any family’s Submit tool. (permissions from AdvancedAnalyticsPermissions.Analytics)
- CancelAdvancedAnalyticsTrainingRun: Cancels a training run submitted by any family’s Submit tool; a finished run is returned unchanged. (permissions from AdvancedAnalyticsPermissions.Training)
- GetAdvancedAnalyticsTrainingRunResult: Reads the outcome of a finished training run: its final status and, once succeeded, the trained model and version plus family-specific evidence. (permissions from AdvancedAnalyticsPermissions.Analytics)
Downstream Dependencies: none (leaf)

Advanced Analytics Training Classifier Margin (`advanced-analytics-training-classifier-margin`)

Hierarchy Placement: Level 4, child of Advanced Analytics Training Supervisor
Semantic Persona: Trains the margin, instance and connectionist classifiers of advanced analytics: linear and kernel support vector machines, k nearest neighbours and a neural network. None emits a probability of its own, so every one relies on calibration fitted on the validation split.
Semantic Routing Keywords: support vector machine, svm, kernel, k nearest neighbours, neural network, classifier, margin
Toolbox (Max 15):
- SubmitLinearSupportVectorClassifierTraining: Queues a linear support vector machine classifier training run, fitting a maximum-margin linear boundary by dual coordinate descent and Platt-calibrating its signed distance. ( permissions from AdvancedAnalyticsPermissions.Training)
- SubmitKernelSupportVectorClassifierTraining: Queues a kernel support vector machine classifier training run, fitting a maximum-margin boundary in a kernel-reshaped space by sequential minimal optimisation and Platt-calibrating its signed distance. (permissions from AdvancedAnalyticsPermissions.Training)
- SubmitKNearestNeighboursClassifierTraining: Queues a k nearest neighbours classifier training run, scoring a row by the Platt-calibrated vote share of its k nearest standardised training rows. (permissions from AdvancedAnalyticsPermissions.Training)
- SubmitNeuralNetworkClassifierTraining: Queues a small feedforward neural network classifier training run, fitting one hidden layer and Platt-calibrating its sigmoid output. (permissions from AdvancedAnalyticsPermissions.Training)
- GetAdvancedAnalyticsTrainingRunStatus: Reads where a queued training run has got to: Pending, Running, then Succeeded, Failed or Cancelled; works for a run submitted by any family’s Submit tool. (permissions from AdvancedAnalyticsPermissions.Analytics)
- CancelAdvancedAnalyticsTrainingRun: Cancels a training run submitted by any family’s Submit tool; a finished run is returned unchanged. (permissions from AdvancedAnalyticsPermissions.Training)
- GetAdvancedAnalyticsTrainingRunResult: Reads the outcome of a finished training run: its final status and, once succeeded, the trained model and version plus family-specific evidence. (permissions from AdvancedAnalyticsPermissions.Analytics)
Downstream Dependencies: none (leaf)

Advanced Analytics Training Clusterer Centroid (`advanced-analytics-training-clusterer-centroid`)

Hierarchy Placement: Level 4, child of Advanced Analytics Training Supervisor
Semantic Persona: Trains the centroid and medoid clusterers of advanced analytics: k means, mini batch k means, balanced k means, k medoids and k modes. Every one is told how many groups to find and summarises each by one representative point.
Semantic Routing Keywords: k means, mini batch k means, balanced k means, k medoids, k modes, clusterer, centroid, medoid
Toolbox (Max 15):
- SubmitKMeansClustererTraining: Queues a k-means clustering training run, assigning rows to the nearest of k iteratively-recentred means with no labels. (permissions from AdvancedAnalyticsPermissions.Training)
- SubmitMiniBatchKMeansClustererTraining: Queues a mini-batch k-means clustering training run: a faster k-means approximation for large row counts that nudges centres from small random batches. (permissions from AdvancedAnalyticsPermissions.Training)
- SubmitBalancedKMeansClustererTraining: Queues a balanced k-means clustering training run, constraining cluster sizes to as close to equal as the row count allows via an optimal-assignment step. (permissions from AdvancedAnalyticsPermissions.Training)
- SubmitKMedoidsClustererTraining: Queues a k-medoids clustering training run, requiring each cluster centre to be an actual training row rather than an averaged point. (permissions from AdvancedAnalyticsPermissions.Training)
- SubmitKModesClustererTraining: Queues a k-modes clustering training run over categorical, integer-coded columns, clustering by per-cluster modal category rather than an averaged centre. (permissions from AdvancedAnalyticsPermissions.Training)
- GetAdvancedAnalyticsTrainingRunStatus: Reads where a queued training run has got to: Pending, Running, then Succeeded, Failed or Cancelled; works for a run submitted by any family’s Submit tool. (permissions from AdvancedAnalyticsPermissions.Analytics)
- CancelAdvancedAnalyticsTrainingRun: Cancels a training run submitted by any family’s Submit tool; a finished run is returned unchanged. (permissions from AdvancedAnalyticsPermissions.Training)
- GetAdvancedAnalyticsTrainingRunResult: Reads the outcome of a finished training run: its final status and, once succeeded, the trained model and version plus family-specific evidence. (permissions from AdvancedAnalyticsPermissions.Analytics)
Downstream Dependencies: none (leaf)

Advanced Analytics Training Clusterer Density (`advanced-analytics-training-clusterer-density`)

Hierarchy Placement: Level 4, child of Advanced Analytics Training Supervisor
Semantic Persona: Trains the hierarchical, density and mixture clusterers of advanced analytics: binary split, mean shift and a Gaussian mixture. These discover how many groups there are rather than being told.
Semantic Routing Keywords: binary split, mean shift, gaussian mixture, clusterer, density, hierarchical
Toolbox (Max 15):
- SubmitBinarySplitClustererTraining: Queues a binary split clustering training run, repeatedly bisecting whichever existing cluster has the largest distortion until k clusters exist. (permissions from AdvancedAnalyticsPermissions.Training)
- SubmitMeanShiftClustererTraining: Queues a mean shift clustering training run, letting rows climb to the nearest density peak so the cluster count is discovered rather than chosen. (permissions from AdvancedAnalyticsPermissions.Training)
- SubmitGaussianMixtureClustererTraining: Queues a Gaussian mixture clustering training run, fitting overlapping bell-shaped components and reporting each row’s per-component membership probability. (permissions from AdvancedAnalyticsPermissions.Training)
- GetAdvancedAnalyticsTrainingRunStatus: Reads where a queued training run has got to: Pending, Running, then Succeeded, Failed or Cancelled; works for a run submitted by any family’s Submit tool. (permissions from AdvancedAnalyticsPermissions.Analytics)
- CancelAdvancedAnalyticsTrainingRun: Cancels a training run submitted by any family’s Submit tool; a finished run is returned unchanged. (permissions from AdvancedAnalyticsPermissions.Training)
- GetAdvancedAnalyticsTrainingRunResult: Reads the outcome of a finished training run: its final status and, once succeeded, the trained model and version plus family-specific evidence. (permissions from AdvancedAnalyticsPermissions.Analytics)
Downstream Dependencies: none (leaf)

Advanced Analytics Training Regressor (`advanced-analytics-training-regressor`)

Hierarchy Placement: Level 4, child of Advanced Analytics Training Supervisor
Semantic Persona: Trains every regressor in advanced analytics: linear regression, kernel ridge, a decision tree, a random forest, gradient boosting, k nearest neighbours and a neural network. All seven predict a number rather than a class or a group.
Semantic Routing Keywords: linear regression, kernel ridge, decision tree, random forest, gradient boosting, k nearest neighbours, neural network, regressor
Toolbox (Max 15):
- SubmitLinearRegressionTraining: Queues an ordinary least squares linear regression training run against inline rows. (permissions from AdvancedAnalyticsPermissions.Training)
- SubmitKernelRidgeRegressorTraining: Queues a kernel ridge regression training run, solving a ridge-penalised fit in a kernel-reshaped space by a single closed-form linear solve. (permissions from AdvancedAnalyticsPermissions.Training)
- SubmitDecisionTreeRegressorTraining: Queues a decision tree regression training run, fitting nested threshold splits that most reduce the outcome’s variance. (permissions from AdvancedAnalyticsPermissions.Training)
- SubmitRandomForestRegressorTraining: Queues a random forest regression training run, averaging many bootstrap-sampled regression trees into a steadier continuous prediction. (permissions from AdvancedAnalyticsPermissions.Training)
- SubmitGradientBoostingRegressorTraining: Queues a gradient boosting regression training run, fitting shallow trees successively to the current residual with no calibration needed. (permissions from AdvancedAnalyticsPermissions.Training)
- SubmitKNearestNeighboursRegressorTraining: Queues a k nearest neighbours regression training run, predicting the mean target of a row’s k nearest standardised training rows. (permissions from AdvancedAnalyticsPermissions.Training)
- SubmitNeuralNetworkRegressorTraining: Queues a small feedforward neural network regression training run, mapping its bounded output back onto the outcome’s original scale. (permissions from AdvancedAnalyticsPermissions.Training)
- GetAdvancedAnalyticsTrainingRunStatus: Reads where a queued training run has got to: Pending, Running, then Succeeded, Failed or Cancelled; works for a run submitted by any family’s Submit tool. (permissions from AdvancedAnalyticsPermissions.Analytics)
- CancelAdvancedAnalyticsTrainingRun: Cancels a training run submitted by any family’s Submit tool; a finished run is returned unchanged. (permissions from AdvancedAnalyticsPermissions.Training)
- GetAdvancedAnalyticsTrainingRunResult: Reads the outcome of a finished training run: its final status and, once succeeded, the trained model and version plus family-specific evidence. (permissions from AdvancedAnalyticsPermissions.Analytics)
Downstream Dependencies: none (leaf)

Advanced Analytics Training Anomaly (`advanced-analytics-training-anomaly`)

Hierarchy Placement: Level 4, child of Advanced Analytics Training Supervisor
Semantic Persona: Trains the anomaly detectors of advanced analytics: isolation forest, Mahalanobis distance and a one-class support vector machine. All three learn what normal looks like and score distance from it with no label required.
Semantic Routing Keywords: isolation forest, mahalanobis, one-class support vector machine, anomaly, outlier
Toolbox (Max 15):
- SubmitIsolationForestAnomalyTraining: Queues an isolation forest anomaly training run, scoring a row by how few random splits it takes to isolate it; unsupervised, with labels only optionally calibrating the score afterwards. (permissions from AdvancedAnalyticsPermissions.Training)
- SubmitMahalanobisAnomalyTraining: Queues a Mahalanobis-distance anomaly training run, scoring a row by its correlation-aware distance from the fitted mean. (permissions from AdvancedAnalyticsPermissions.Training)
- SubmitOneClassSupportVectorAnomalyTraining: Queues a one-class support vector machine anomaly training run, fitting a boundary that wraps snugly around the bulk of unlabelled rows in a kernel-reshaped space. (permissions from AdvancedAnalyticsPermissions.Training)
- GetAdvancedAnalyticsTrainingRunStatus: Reads where a queued training run has got to: Pending, Running, then Succeeded, Failed or Cancelled; works for a run submitted by any family’s Submit tool. (permissions from AdvancedAnalyticsPermissions.Analytics)
- CancelAdvancedAnalyticsTrainingRun: Cancels a training run submitted by any family’s Submit tool; a finished run is returned unchanged. (permissions from AdvancedAnalyticsPermissions.Training)
- GetAdvancedAnalyticsTrainingRunResult: Reads the outcome of a finished training run: its final status and, once succeeded, the trained model and version plus family-specific evidence. (permissions from AdvancedAnalyticsPermissions.Analytics)
Downstream Dependencies: none (leaf)

Advanced Analytics Training Latent Structure (`advanced-analytics-training-latent-structure`)

Hierarchy Placement: Level 4, child of Advanced Analytics Training Supervisor
Semantic Persona: Trains the two families of advanced analytics that describe the shape of the data rather than predicting from it: principal components, which reduces dimension, and a Gaussian mixture density, which estimates the distribution itself. Neither predicts anything.
Semantic Routing Keywords: principal component, pca, gaussian mixture density, dimension reduction, density estimation, latent structure
Toolbox (Max 15):
- SubmitPrincipalComponentTraining: Queues a principal component training run, re-expressing correlated variables as a smaller set of uncorrelated, variance-ranked components. (permissions from AdvancedAnalyticsPermissions.Training)
- SubmitGaussianMixtureDensityTraining: Queues a Gaussian mixture density training run, describing the rows’ shape as a weighted sum of components and scoring a row by its log-density. (permissions from AdvancedAnalyticsPermissions.Training)
- GetAdvancedAnalyticsTrainingRunStatus: Reads where a queued training run has got to: Pending, Running, then Succeeded, Failed or Cancelled; works for a run submitted by any family’s Submit tool. (permissions from AdvancedAnalyticsPermissions.Analytics)
- CancelAdvancedAnalyticsTrainingRun: Cancels a training run submitted by any family’s Submit tool; a finished run is returned unchanged. (permissions from AdvancedAnalyticsPermissions.Training)
- GetAdvancedAnalyticsTrainingRunResult: Reads the outcome of a finished training run: its final status and, once succeeded, the trained model and version plus family-specific evidence. (permissions from AdvancedAnalyticsPermissions.Analytics)
Downstream Dependencies: none (leaf)

Exhaustive Supervisor

Hierarchy Placement: Level 3, child of Model Risk and Analytics
Semantic Persona: Routes within the exhaustive slice of model risk and analytics. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: exhaustive, search, instance, trial, list, query, variable, promoted, confusion, error, histogram, learning, curve, predicted, actual, roc, prescription, promotion, history, variance, statistics
Toolbox (Max 15): none — routes to 6 leaf agents covering 28 tools
Downstream Dependencies: exhaustive-search-instance-inspector, exhaustive-search-instance-test-bench, exhaustive-search-instance-author, exhaustive-trials-and-confusion, exhaustive-promotion-curves-and-history, exhaustive-variable-statistics
Cross-Domain Reach: completions

Exhaustive Search Instance Inspector (`exhaustive-search-instance-inspector`)

Hierarchy Placement: Level 4, child of Exhaustive Supervisor
Semantic Persona: Owns the exhaustive search instance slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: exhaustive, search, instance, inspector
Toolbox (Max 15):
- ExhaustiveSearchInstanceList: Lists Exhaustive Adaptations for the caller’s tenant, keyset-paged and capped. ( permissions from ExhaustiveSearchInstanceService.ListPermissions)
- ExhaustiveSearchInstanceFilterFields: Lists the fields a query builder JSON filter over Exhaustive Search Instances may use. (permissions from ExhaustiveSearchInstanceService.ListPermissions)
- ExhaustiveSearchInstanceFilter: Returns the Exhaustive Search Instances matching query builder JSON, keyset-paged and capped. (permissions from ExhaustiveSearchInstanceService.ListPermissions)
- ExhaustiveSearchInstanceCount: Counts the Exhaustive Search Instances matching query builder JSON, optionally grouped by a field. (permissions from ExhaustiveSearchInstanceService.ListPermissions)
- ExhaustiveSearchInstanceGet: Returns one Exhaustive Adaptation by id, scoped to the caller’s tenant. ( permissions from ExhaustiveSearchInstanceService.ReadPermissions)
- ExhaustiveSearchInstanceGetByGuid: Returns one Exhaustive Adaptation by Guid, scoped to the caller’s tenant. ( permissions from ExhaustiveSearchInstanceService.ReadPermissions)
- ExhaustiveSearchInstanceGetByEntityAnalysisModelId: Lists Exhaustive Adaptations belonging to a given Model, scoped to the caller’s tenant. (permissions from ExhaustiveSearchInstanceService.ReadPermissions)
Downstream Dependencies: none (leaf)

Exhaustive Search Instance Test Bench (`exhaustive-search-instance-test-bench`)

Hierarchy Placement: Level 4, child of Exhaustive Supervisor
Semantic Persona: Owns the exhaustive search instance slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: exhaustive, search, instance, test, bench
Toolbox (Max 15):
- ExhaustiveSearchInstanceValidate: Validates an Exhaustive Search Instance without saving it and returns every failure. (permissions from ExhaustiveSearchInstanceService.WritePermissions)
Downstream Dependencies: none (leaf)

Exhaustive Search Instance Author (`exhaustive-search-instance-author`)

Hierarchy Placement: Level 4, child of Exhaustive Supervisor
Semantic Persona: Owns the exhaustive search instance slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: exhaustive, search, instance, author
Toolbox (Max 15):
- ExhaustiveSearchInstanceList: Lists Exhaustive Adaptations for the caller’s tenant, keyset-paged and capped. ( permissions from ExhaustiveSearchInstanceService.ListPermissions)
- ExhaustiveSearchInstanceFilterFields: Lists the fields a query builder JSON filter over Exhaustive Search Instances may use. (permissions from ExhaustiveSearchInstanceService.ListPermissions)
- ExhaustiveSearchInstanceFilter: Returns the Exhaustive Search Instances matching query builder JSON, keyset-paged and capped. (permissions from ExhaustiveSearchInstanceService.ListPermissions)
- ExhaustiveSearchInstanceCount: Counts the Exhaustive Search Instances matching query builder JSON, optionally grouped by a field. (permissions from ExhaustiveSearchInstanceService.ListPermissions)
- ExhaustiveSearchInstanceGet: Returns one Exhaustive Adaptation by id, scoped to the caller’s tenant. ( permissions from ExhaustiveSearchInstanceService.ReadPermissions)
- ExhaustiveSearchInstanceGetByGuid: Returns one Exhaustive Adaptation by Guid, scoped to the caller’s tenant. ( permissions from ExhaustiveSearchInstanceService.ReadPermissions)
- ExhaustiveSearchInstanceGetByEntityAnalysisModelId: Lists Exhaustive Adaptations belonging to a given Model, scoped to the caller’s tenant. (permissions from ExhaustiveSearchInstanceService.ReadPermissions)
- ExhaustiveSearchInstanceValidate: Validates an Exhaustive Search Instance without saving it and returns every failure. (permissions from ExhaustiveSearchInstanceService.WritePermissions)
- ExhaustiveSearchInstanceCreate: Registers an Exhaustive Adaptation under a Model in the caller’s tenant; calling twice creates two. Write — confirmation required (permissions from ExhaustiveSearchInstanceService.WritePermissions)
- ExhaustiveSearchInstanceUpdate: Updates an Exhaustive Adaptation in the caller’s tenant by id, only while training has not started. Write — confirmation required (permissions from ExhaustiveSearchInstanceService.WritePermissions)
- ExhaustiveSearchInstanceStop: Requests that an in-progress or queued training run stop, by Guid. Write — confirmation required (permissions from ExhaustiveSearchInstanceService.WritePermissions)
- ExhaustiveSearchInstanceDelete: Soft-deletes an Exhaustive Adaptation in the caller’s tenant by id. Write — confirmation required (permissions from ExhaustiveSearchInstanceService.WritePermissions)
- ExhaustiveSearchInstanceSetActive: Sets whether an Exhaustive Adaptation is Active in the caller’s tenant. Write — confirmation required (permissions from ExhaustiveSearchInstanceService.WritePermissions)
- ExhaustiveSearchInstanceResubmit: Resubmits a rejected Exhaustive Adaptation for approval again, unchanged, opening a fresh approval cycle. Write — confirmation required (permissions from ExhaustiveSearchInstanceService.WritePermissions)
- ExhaustiveSearchInstanceRetract: Retracts a rejected Exhaustive Adaptation, restoring it to its last approved version without a new approval cycle. Write — confirmation required (permissions from ExhaustiveSearchInstanceService.WritePermissions)
Downstream Dependencies: none (leaf)

Exhaustive Trials and Confusion (`exhaustive-trials-and-confusion`)

Hierarchy Placement: Level 4, child of Exhaustive Supervisor
Semantic Persona: Owns the exhaustive trials and confusion slice end to end, which is small enough that reads and writes stay with one agent. It holds 2 mutating tools, so a confirmation is required before any of them runs.
Semantic Routing Keywords: exhaustive, search, instance, trial, list, query, variable, promoted, confusion, error, histogram
Toolbox (Max 15):
- ExhaustiveSearchInstanceTrialInstanceListQueryGet: Lists every trial attempt of one Exhaustive Search Instance, scoped to the caller’s tenant. (permissions from ExhaustiveSearchInstanceTrialInstanceListQueryService.RequiredPermissions)
- ExhaustiveSearchInstanceTrialInstanceDetailGet: Returns the detail of one Exhaustive trial attempt, scoped to the caller’s tenant. (permissions from ExhaustiveSearchInstanceTrialInstanceListQueryService.RequiredPermissions)
- ExhaustiveSearchInstanceVariableGet: Lists the variables of an Exhaustive Search Instance with distribution statistics and histogram bins. (permissions from ExhaustiveSearchInstanceVariableService.RequiredPermissions)
- ExhaustiveSearchInstanceVariableReallocate: Reallocates a trained variable’s live name and processing type, without retraining. Write — confirmation required (permissions from ExhaustiveSearchInstanceVariableService.RequiredPermissions)
- ExhaustiveSearchInstancePromotedTrialInstanceUpdate: Sets whether a promoted trial instance is promoted for production use; setting the same value twice has no further effect. Write — confirmation required ( permissions from ExhaustiveSearchInstancePromotedTrialInstanceService.WritePermissions)
- ExhaustiveSearchInstancePromotedTrialInstanceConfusionGet: Returns the confusion matrix and derived ratios for the active promoted trial instance of an Exhaustive Search Instance. (permissions from ExhaustiveSearchInstancePromotedTrialInstanceConfusionService.RequiredPermissions)
- ExhaustiveSearchInstancePromotedTrialInstanceErrorHistogramGet: Returns a ten-bin histogram of prediction errors of the active promoted trial instance of an exhaustive search instance. (permissions from ExhaustiveSearchInstancePromotedTrialInstanceErrorHistogramService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Exhaustive Promotion Curves and History (`exhaustive-promotion-curves-and-history`)

Hierarchy Placement: Level 4, child of Exhaustive Supervisor
Semantic Persona: Owns the exhaustive promotion curves and history slice end to end, which is small enough that reads and writes stay with one agent.
Semantic Routing Keywords: exhaustive, search, instance, promoted, trial, learning, curve, predicted, actual, query, roc, variable, prescription, promotion, history
Toolbox (Max 15):
- ExhaustiveSearchInstancePromotedTrialInstanceLearningCurveGet: Returns the score-over-time learning curve of an exhaustive search instance’s promoted trial instances. (permissions from ExhaustiveSearchInstancePromotedTrialInstanceLearningCurveService.RequiredPermissions)
- ExhaustiveSearchInstancePromotedTrialInstancePredictedActualGet: Returns predicted, actual and error values for the promoted trial instance of an exhaustive search instance. (permissions from ExhaustiveSearchInstancePromotedTrialInstancePredictedActualService.RequiredPermissions)
- ExhaustiveSearchInstancePromotedTrialInstanceQueryGet: Lists the promoted trial instances of one Exhaustive Search Instance, newest first. (permissions from ExhaustiveSearchInstancePromotedTrialInstanceQueryService.RequiredPermissions)
- ExhaustiveSearchInstancePromotedTrialInstanceRocGet: Returns the ROC curve points (score, false positive rate, true positive rate) of an Exhaustive Search Instance’s active promoted trial instance. (permissions from ExhaustiveSearchInstancePromotedTrialInstanceRocService.RequiredPermissions)
- ExhaustiveSearchInstancePromotedTrialInstanceVariablePrescriptionGet: Lists the variables of a promoted trial instance with variable and prescription statistics and sensitivity, ordered by sensitivity descending. ( permissions from ExhaustiveSearchInstancePromotedTrialInstanceVariablePrescriptionService.RequiredPermissions)
- ExhaustiveSearchInstancePromotionHistoryGet: Returns the promotion history of an exhaustive search instance: score, topology complexity and trials explored at each promotion. (permissions from ExhaustiveSearchInstancePromotionHistoryService.RequiredPermissions)
- ExhaustiveSearchInstanceTrialInstanceVariableGet: Lists the variables of the promoted trial instance of an Exhaustive Search Instance with normalisation statistics, ordered by variable sequence. (permissions from ExhaustiveSearchInstanceTrialInstanceVariableService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Exhaustive Variable Statistics (`exhaustive-variable-statistics`)

Hierarchy Placement: Level 4, child of Exhaustive Supervisor
Semantic Persona: Owns the exhaustive variable statistics slice end to end, which is small enough that reads and writes stay with one agent.
Semantic Routing Keywords: exhaustive, search, instance, trial, variable, variance, statistics
Toolbox (Max 15):
- ExhaustiveSearchInstanceTrialInstanceVariableVarianceGet: Lists the multi-collinearity correlations of one Exhaustive Search Instance Variable against its test variables, ordered by absolute correlation rank. ( permissions from ExhaustiveSearchInstanceTrialInstanceVariableVarianceService.RequiredPermissions)
- ExhaustiveSearchInstanceVariableStatisticsGet: Lists the variables of an Exhaustive Search Instance, or one variable, with the population, positive class, negative class and Monte Carlo prescription histograms. ( permissions from ExhaustiveSearchInstanceVariableStatisticsService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Validation Supervisor

Hierarchy Placement: Level 3, child of Model Risk and Analytics
Semantic Persona: Routes within the validation slice of model risk and analytics. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: validation, entity, analysis, model, backtest, http, adaptation, integrity, history, simulation, sample
Toolbox (Max 15): none — routes to 7 leaf agents covering 25 tools
Downstream Dependencies: entity-analysis-model-backtest-inspector, entity-analysis-model-backtest-test-bench, entity-analysis-model-backtest-author, entity-analysis-model-http-adaptation-inspector, entity-analysis-model-http-adaptation-test-bench, entity-analysis-model-http-adaptation-author, model-integrity-and-sampling
Cross-Domain Reach: completions

Entity Analysis Model Backtest Inspector (`entity-analysis-model-backtest-inspector`)

Hierarchy Placement: Level 4, child of Validation Supervisor
Semantic Persona: Owns the entity analysis model backtest slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: entity, analysis, model, backtest, inspector
Toolbox (Max 15):
- EntityAnalysisModelBacktestFields: Lists the fields a backtest’s filter and class may use, including Tag. for the class. _(permissions from `EntityAnalysisModelBacktestService.RequiredPermissions`)_
- EntityAnalysisModelBacktestRunRule: Backtests a rule immediately over a filtered archive against a class; returns raw counts. (permissions from EntityAnalysisModelBacktestService.RequiredPermissions)
- EntityAnalysisModelBacktestStatus: Returns a submitted backtest’s status, progress and headline counts. ( permissions from EntityAnalysisModelBacktestService.RequiredPermissions)
- EntityAnalysisModelBacktestList: Lists a model’s submitted backtests, optionally for one rule. (permissions from EntityAnalysisModelBacktestService.RequiredPermissions)
- EntityAnalysisModelBacktestResult: Returns a finished backtest’s counts and examples. (permissions from EntityAnalysisModelBacktestService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Backtest Test Bench (`entity-analysis-model-backtest-test-bench`)

Hierarchy Placement: Level 4, child of Validation Supervisor
Semantic Persona: Owns the entity analysis model backtest slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: entity, analysis, model, backtest, test, bench
Toolbox (Max 15):
- EntityAnalysisModelBacktestExplain: Explains why one archived transaction was or was not counted, with the values read. (permissions from EntityAnalysisModelBacktestService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Backtest Author (`entity-analysis-model-backtest-author`)

Hierarchy Placement: Level 4, child of Validation Supervisor
Semantic Persona: Owns the entity analysis model backtest slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: entity, analysis, model, backtest, author
Toolbox (Max 15):
- EntityAnalysisModelBacktestFields: Lists the fields a backtest’s filter and class may use, including Tag. for the class. _(permissions from `EntityAnalysisModelBacktestService.RequiredPermissions`)_
- EntityAnalysisModelBacktestRunRule: Backtests a rule immediately over a filtered archive against a class; returns raw counts. (permissions from EntityAnalysisModelBacktestService.RequiredPermissions)
- EntityAnalysisModelBacktestStatus: Returns a submitted backtest’s status, progress and headline counts. ( permissions from EntityAnalysisModelBacktestService.RequiredPermissions)
- EntityAnalysisModelBacktestList: Lists a model’s submitted backtests, optionally for one rule. (permissions from EntityAnalysisModelBacktestService.RequiredPermissions)
- EntityAnalysisModelBacktestResult: Returns a finished backtest’s counts and examples. (permissions from EntityAnalysisModelBacktestService.RequiredPermissions)
- EntityAnalysisModelBacktestExplain: Explains why one archived transaction was or was not counted, with the values read. (permissions from EntityAnalysisModelBacktestService.RequiredPermissions)
- EntityAnalysisModelBacktestSubmit: Submits a large backtest to the engine’s backtest threads; returns the run to poll. Write — confirmation required (permissions from EntityAnalysisModelBacktestService.RequiredPermissions)
- EntityAnalysisModelBacktestStop: Stops a submitted backtest. Write — confirmation required (permissions from EntityAnalysisModelBacktestService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Http Adaptation Inspector (`entity-analysis-model-http-adaptation-inspector`)

Hierarchy Placement: Level 4, child of Validation Supervisor
Semantic Persona: Owns the entity analysis model http adaptation slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: entity, analysis, model, http, adaptation, inspector
Toolbox (Max 15):
- EntityAnalysisModelHttpAdaptationList: Lists HTTP Adaptations for the caller’s tenant, keyset-paged and capped. (permissions from EntityAnalysisModelHttpAdaptationService.ListPermissions)
- EntityAnalysisModelHttpAdaptationFilterFields: Lists the fields a query builder JSON filter over HTTP Adaptations may use. (permissions from EntityAnalysisModelHttpAdaptationService.ListPermissions)
- EntityAnalysisModelHttpAdaptationFilter: Returns the HTTP Adaptations matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelHttpAdaptationService.ListPermissions)
- EntityAnalysisModelHttpAdaptationCount: Counts the HTTP Adaptations matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelHttpAdaptationService.ListPermissions)
- EntityAnalysisModelHttpAdaptationGet: Returns one HTTP Adaptation by id, scoped to the caller’s tenant. ( permissions from EntityAnalysisModelHttpAdaptationService.ReadPermissions)
- EntityAnalysisModelHttpAdaptationGetByGuid: Returns one HTTP Adaptation by Guid, scoped to the caller’s tenant. (permissions from EntityAnalysisModelHttpAdaptationService.ReadPermissions)
- EntityAnalysisModelHttpAdaptationGetByEntityAnalysisModelId: Lists HTTP Adaptations belonging to a given Model, scoped to the caller’s tenant. (permissions from EntityAnalysisModelHttpAdaptationService.ReadPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Http Adaptation Test Bench (`entity-analysis-model-http-adaptation-test-bench`)

Hierarchy Placement: Level 4, child of Validation Supervisor
Semantic Persona: Owns the entity analysis model http adaptation slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: entity, analysis, model, http, adaptation, test, bench
Toolbox (Max 15):
- EntityAnalysisModelHttpAdaptationValidate: Validates a HTTP Adaptation without saving it and returns every failure. (permissions from EntityAnalysisModelHttpAdaptationService.WritePermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Http Adaptation Author (`entity-analysis-model-http-adaptation-author`)

Hierarchy Placement: Level 4, child of Validation Supervisor
Semantic Persona: Owns the entity analysis model http adaptation slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: entity, analysis, model, http, adaptation, author
Toolbox (Max 15):
- EntityAnalysisModelHttpAdaptationList: Lists HTTP Adaptations for the caller’s tenant, keyset-paged and capped. (permissions from EntityAnalysisModelHttpAdaptationService.ListPermissions)
- EntityAnalysisModelHttpAdaptationFilterFields: Lists the fields a query builder JSON filter over HTTP Adaptations may use. (permissions from EntityAnalysisModelHttpAdaptationService.ListPermissions)
- EntityAnalysisModelHttpAdaptationFilter: Returns the HTTP Adaptations matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelHttpAdaptationService.ListPermissions)
- EntityAnalysisModelHttpAdaptationCount: Counts the HTTP Adaptations matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelHttpAdaptationService.ListPermissions)
- EntityAnalysisModelHttpAdaptationGet: Returns one HTTP Adaptation by id, scoped to the caller’s tenant. ( permissions from EntityAnalysisModelHttpAdaptationService.ReadPermissions)
- EntityAnalysisModelHttpAdaptationGetByGuid: Returns one HTTP Adaptation by Guid, scoped to the caller’s tenant. (permissions from EntityAnalysisModelHttpAdaptationService.ReadPermissions)
- EntityAnalysisModelHttpAdaptationGetByEntityAnalysisModelId: Lists HTTP Adaptations belonging to a given Model, scoped to the caller’s tenant. (permissions from EntityAnalysisModelHttpAdaptationService.ReadPermissions)
- EntityAnalysisModelHttpAdaptationValidate: Validates a HTTP Adaptation without saving it and returns every failure. (permissions from EntityAnalysisModelHttpAdaptationService.WritePermissions)
- EntityAnalysisModelHttpAdaptationCreate: Registers an HTTP Adaptation under a Model in the caller’s tenant; calling twice creates two. Write — confirmation required (permissions from EntityAnalysisModelHttpAdaptationService.WritePermissions)
- EntityAnalysisModelHttpAdaptationUpdate: Updates an HTTP Adaptation in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelHttpAdaptationService.WritePermissions)
- EntityAnalysisModelHttpAdaptationDelete: Soft-deletes an HTTP Adaptation in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelHttpAdaptationService.WritePermissions)
- EntityAnalysisModelHttpAdaptationResubmit: Resubmits a rejected HTTP Adaptation for approval again, unchanged, opening a fresh approval cycle. Write — confirmation required (permissions from EntityAnalysisModelHttpAdaptationService.WritePermissions)
- EntityAnalysisModelHttpAdaptationRetract: Retracts a rejected HTTP Adaptation, restoring it to its last approved version without a new approval cycle. Write — confirmation required (permissions from EntityAnalysisModelHttpAdaptationService.WritePermissions)
Downstream Dependencies: none (leaf)

Model Integrity and Sampling (`model-integrity-and-sampling`)

Hierarchy Placement: Level 4, child of Validation Supervisor
Semantic Persona: Owns the model integrity and sampling slice end to end, which is small enough that reads and writes stay with one agent.
Semantic Routing Keywords: entity, analysis, model, integrity, history, simulation, sample
Toolbox (Max 15):
- EntityAnalysisModelIntegrityModels: Lists the tenant’s models that can be checked. (permissions from EntityAnalysisModelIntegrityService.RequiredPermissions)
- EntityAnalysisModelIntegrityCheck: Checks a model’s dependencies, compilation, configuration and engine state; returns coded findings. (permissions from EntityAnalysisModelIntegrityService.RequiredPermissions)
- EntityAnalysisModelIntegrityEngineState: Returns what the engine instances have loaded and started for a model. (permissions from EntityAnalysisModelIntegrityService.RequiredPermissions)
- EntityAnalysisModelIntegrityGraph: Returns a model’s dependency graph as nodes and edges, optionally centred on one entity. (permissions from EntityAnalysisModelIntegrityService.RequiredPermissions)
- EntityAnalysisModelHistorySimulationAbstractionAggregation: Works out an abstraction rule’s value for a transaction from cached history, as the engine does. (permissions from EntityAnalysisModelHistorySimulationService.RequiredPermissions)
- EntityAnalysisModelHistorySimulationTtlCounterValue: Reads a TTL counter’s value for a transaction from the cache, as the engine does; increments nothing. (permissions from EntityAnalysisModelHistorySimulationService.RequiredPermissions)
- EntityAnalysisModelSampleExecute: Samples archived transactions of an Entity Analysis Model over a date window and returns them as a CSV file. (permissions from EntityAnalysisModelSampleService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Model Supporting Structures (`model-support`)

Model Supporting Structures

Hierarchy Placement: Level 2, child of Jooby
Semantic Persona: Holds the lookup and operational scaffolding a detection model leans on: lists and dictionaries, overrides, reprocessing, the configuration navigation tree and the authoring aids. Separate from the model builder because these are shared structures with their own lifecycles.
Semantic Routing Keywords: list, dictionary, lookup, override, suppression, reprocessing, replay, tree, completion, parser, rule vocabulary
Toolbox (Max 15): none — routes to 6 supervisors covering 128 tools
Downstream Dependencies: model-support.authoring-support, model-support.data-structures.dictionaries, model-support.data-structures.lists, model-support.navigation, model-support.overrides, model-support.reprocessing
Cross-Domain Reach: entity-analysis-model-synchronisation-schedule

Authoring Support Supervisor

Hierarchy Placement: Level 3, child of Model Supporting Structures
Semantic Persona: Routes within the authoring support slice of model supporting structures. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: authoring support, completions, entity, analysis, model, invocation, context, dependency, role, rule, vocabulary, icons, parser
Toolbox (Max 15): none — routes to 3 leaf agents covering 21 tools
Downstream Dependencies: completions, model-invocation-context-and-dependencies, authoring-aids-and-model-role-grants

Completions (`completions`)

Hierarchy Placement: Level 4, child of Authoring Support Supervisor
Semantic Persona: Owns the completions slice end to end, which is small enough that reads and writes stay with one agent.
Semantic Routing Keywords: completions
Toolbox (Max 15):
- CompletionsGetByCaseWorkflowId: Gets the payload field completions for the model behind a case workflow, by workflow Id. (permissions from CompletionsService.ByCaseWorkflowIdPermissions)
- CompletionsGetByCaseWorkflowGuidIncludingDeleted: Gets the payload field completions for the model behind a case workflow, by workflow Guid, including deleted workflows. (permissions from CompletionsService.IncludingDeletedPermissions)
- CompletionsGetByCaseWorkflowIdIncludingDeleted: Gets the payload field completions for the model behind a case workflow, by workflow Id, including deleted workflows. (permissions from CompletionsService.IncludingDeletedPermissions)
- CompletionsGetByEntityAnalysisModelId: Gets the payload field completions for an entity analysis model, for reporting. (permissions from CompletionsService.ByEntityAnalysisModelIdPermissions)
- CompletionsGetByEntityAnalysisModelIdParseTypeId: Gets the payload field completions for an entity analysis model and parse type. (permissions from CompletionsService.ByParseTypePermissions)
- CompletionsGetByExhaustiveInstanceDataset: Gets filter builder completions for every column of a confirmed uploaded Exhaustive dataset. (permissions from CompletionsService.ByExhaustiveInstanceDatasetPermissions)
- QueryBuilderOperatorsList: Lists every operator the rule builder offers, with the rule text each produces, the values it takes and the field data types it applies to. (permissions from QueryBuilderService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Model Invocation Context and Dependencies (`model-invocation-context-and-dependencies`)

Hierarchy Placement: Level 4, child of Authoring Support Supervisor
Semantic Persona: Owns the model invocation context and dependencies slice end to end, which is small enough that reads and writes stay with one agent.
Semantic Routing Keywords: entity, analysis, model, invocation, context, dependency
Toolbox (Max 15):
- EntityAnalysisModelInvocationContextBlank: Returns a blank rule-evaluation context for a model, with every name a rule can use and its defaults. (permissions from EntityAnalysisModelInvocationContextService.AuthoringPermissions)
- EntityAnalysisModelInvocationContextFromRequestJson: Builds a rule-evaluation context from request JSON with the engine’s own extraction; stores nothing. (permissions from EntityAnalysisModelInvocationContextService.AuthoringPermissions)
- EntityAnalysisModelInvocationContextFromArchive: Builds a rule-evaluation context from an archived transaction, with every value as it was. (permissions from EntityAnalysisModelInvocationContextService.ArchivePermissions)
- EntityAnalysisModelInvocationContextOverlay: Sets values in a rule-evaluation context by completion name, checking each against its data type. (permissions from EntityAnalysisModelInvocationContextService.AuthoringPermissions)
- EntityAnalysisModelDependencyDependents: Lists every entity that uses an entity, directly or indirectly, and whether deleting it would be refused. (permissions from EntityAnalysisModelDependencyService.RequiredPermissions)
- EntityAnalysisModelDependencyDependencies: Lists what an entity uses, resolved to the entities it names. ( permissions from EntityAnalysisModelDependencyService.RequiredPermissions)
- EntityAnalysisModelDependencyIssues: Finds dangling names, uses of inactive entities and unused entities in a model. (permissions from EntityAnalysisModelDependencyService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Authoring Aids and Model Role Grants (`authoring-aids-and-model-role-grants`)

Hierarchy Placement: Level 4, child of Authoring Support Supervisor
Semantic Persona: Owns the authoring aids and model role grants slice end to end, which is small enough that reads and writes stay with one agent. It holds 2 mutating tools, so a confirmation is required before any of them runs.
Semantic Routing Keywords: entity, analysis, model, role, rule, vocabulary, icons, parser
Toolbox (Max 15):
- EntityAnalysisModelRoleListByEntityAnalysisModelGuid: Lists the Role grants for an Entity Analysis Model. ( permissions from EntityAnalysisModelRoleService.RequiredPermissions)
- EntityAnalysisModelRoleCreate: Grants a Role access to an Entity Analysis Model; calling twice creates two grants. Write — confirmation required (permissions from EntityAnalysisModelRoleService.RequiredPermissions)
- EntityAnalysisModelRoleDelete: Removes a Role’s grant on an Entity Analysis Model; reversible only by direct database action. Write — confirmation required (permissions from EntityAnalysisModelRoleService.RequiredPermissions)
- RuleVocabularyList: Lists the keywords, allowed words and rule functions rule text may use. (permissions from RuleVocabularyService.RequiredPermissions)
- RuleVocabularyDescribe: Describes a rule word, with every function signature. (permissions from RuleVocabularyService.RequiredPermissions)
- IconsGet: Lists the icon files available to the user interface. (permissions from IconsService.RequiredPermissions)
- ParserParse: Parses and compiles a user-authored rule against an Entity Analysis Model and returns any errors. (permissions from ParserService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Data Structures Dictionaries Supervisor

Hierarchy Placement: Level 3, child of Model Supporting Structures
Semantic Persona: Routes within the data structures dictionaries slice of model supporting structures. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: data structures dictionaries, entity, analysis, model, dictionary, kvp
Toolbox (Max 15): none — routes to 6 leaf agents covering 21 tools
Downstream Dependencies: entity-analysis-model-dictionary-inspector, entity-analysis-model-dictionary-test-bench, entity-analysis-model-dictionary-author, entity-analysis-model-dictionary-kvp-inspector, entity-analysis-model-dictionary-kvp-test-bench, entity-analysis-model-dictionary-kvp-author

Entity Analysis Model Dictionary Inspector (`entity-analysis-model-dictionary-inspector`)

Hierarchy Placement: Level 4, child of Data Structures Dictionaries Supervisor
Semantic Persona: Owns the entity analysis model dictionary slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: entity, analysis, model, dictionary, inspector
Toolbox (Max 15):
- EntityAnalysisModelDictionaryList: Lists Dictionaries for the caller’s tenant, keyset-paged and capped. ( permissions from EntityAnalysisModelDictionaryService.RequiredPermissions)
- EntityAnalysisModelDictionaryFilterFields: Lists the fields a query builder JSON filter over Dictionarys may use. (permissions from EntityAnalysisModelDictionaryService.RequiredPermissions)
- EntityAnalysisModelDictionaryFilter: Returns the Dictionarys matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelDictionaryService.RequiredPermissions)
- EntityAnalysisModelDictionaryCount: Counts the Dictionarys matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelDictionaryService.RequiredPermissions)
- EntityAnalysisModelDictionaryGet: Returns one Dictionary by id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelDictionaryService.RequiredPermissions)
- EntityAnalysisModelDictionaryGetByGuid: Returns one Dictionary by Guid, scoped to the caller’s tenant. ( permissions from EntityAnalysisModelDictionaryService.RequiredPermissions)
- EntityAnalysisModelDictionaryGetByEntityAnalysisModelId: Lists the Dictionaries belonging to a given Model, ordered by Id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelDictionaryService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Dictionary Test Bench (`entity-analysis-model-dictionary-test-bench`)

Hierarchy Placement: Level 4, child of Data Structures Dictionaries Supervisor
Semantic Persona: Owns the entity analysis model dictionary slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: entity, analysis, model, dictionary, test, bench
Toolbox (Max 15):
- EntityAnalysisModelDictionaryValidate: Validates a Dictionary without saving it and returns every failure. ( permissions from EntityAnalysisModelDictionaryService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Dictionary Author (`entity-analysis-model-dictionary-author`)

Hierarchy Placement: Level 4, child of Data Structures Dictionaries Supervisor
Semantic Persona: Owns the entity analysis model dictionary slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: entity, analysis, model, dictionary, author
Toolbox (Max 15):
- EntityAnalysisModelDictionaryList: Lists Dictionaries for the caller’s tenant, keyset-paged and capped. ( permissions from EntityAnalysisModelDictionaryService.RequiredPermissions)
- EntityAnalysisModelDictionaryFilterFields: Lists the fields a query builder JSON filter over Dictionarys may use. (permissions from EntityAnalysisModelDictionaryService.RequiredPermissions)
- EntityAnalysisModelDictionaryFilter: Returns the Dictionarys matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelDictionaryService.RequiredPermissions)
- EntityAnalysisModelDictionaryCount: Counts the Dictionarys matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelDictionaryService.RequiredPermissions)
- EntityAnalysisModelDictionaryGet: Returns one Dictionary by id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelDictionaryService.RequiredPermissions)
- EntityAnalysisModelDictionaryGetByGuid: Returns one Dictionary by Guid, scoped to the caller’s tenant. ( permissions from EntityAnalysisModelDictionaryService.RequiredPermissions)
- EntityAnalysisModelDictionaryGetByEntityAnalysisModelId: Lists the Dictionaries belonging to a given Model, ordered by Id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelDictionaryService.RequiredPermissions)
- EntityAnalysisModelDictionaryValidate: Validates a Dictionary without saving it and returns every failure. ( permissions from EntityAnalysisModelDictionaryService.RequiredPermissions)
- EntityAnalysisModelDictionaryCreate: Registers a Dictionary under a Model in the caller’s tenant; calling twice creates two. Write — confirmation required (permissions from EntityAnalysisModelDictionaryService.RequiredPermissions)
- EntityAnalysisModelDictionaryUpdate: Updates a Dictionary in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelDictionaryService.RequiredPermissions)
- EntityAnalysisModelDictionaryDelete: Soft-deletes a Dictionary in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelDictionaryService.RequiredPermissions)
- EntityAnalysisModelDictionaryResubmit: Resubmits a rejected Dictionary for approval again, unchanged, opening a fresh approval cycle. Write — confirmation required (permissions from EntityAnalysisModelDictionaryService.RequiredPermissions)
- EntityAnalysisModelDictionaryRetract: Retracts a rejected Dictionary, restoring it to its last approved version without a new approval cycle. Write — confirmation required (permissions from EntityAnalysisModelDictionaryService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Dictionary Kvp Inspector (`entity-analysis-model-dictionary-kvp-inspector`)

Hierarchy Placement: Level 4, child of Data Structures Dictionaries Supervisor
Semantic Persona: Owns the entity analysis model dictionary kvp slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: entity, analysis, model, dictionary, kvp, inspector
Toolbox (Max 15):
- EntityAnalysisModelDictionaryKvpList: Lists Key Value Pairs for the caller’s tenant, keyset-paged and capped. ( permissions from EntityAnalysisModelDictionaryKvpService.RequiredPermissions)
- EntityAnalysisModelDictionaryKvpFilterFields: Lists the fields a query builder JSON filter over Dictionary KVPs may use. (permissions from EntityAnalysisModelDictionaryKvpService.RequiredPermissions)
- EntityAnalysisModelDictionaryKvpFilter: Returns the Dictionary KVPs matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelDictionaryKvpService.RequiredPermissions)
- EntityAnalysisModelDictionaryKvpCount: Counts the Dictionary KVPs matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelDictionaryKvpService.RequiredPermissions)
- EntityAnalysisModelDictionaryKvpGet: Returns one Key Value Pair, matched on the parent Dictionary’s id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelDictionaryKvpService.RequiredPermissions)
- EntityAnalysisModelDictionaryKvpGetByGuid: Returns one Key Value Pair, matched on the parent Dictionary’s id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelDictionaryKvpService.RequiredPermissions)
- EntityAnalysisModelDictionaryKvpGetByEntityAnalysisModelDictionaryId: Lists the Key Value Pairs belonging to a given Dictionary, ordered by Id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelDictionaryKvpService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Dictionary Kvp Test Bench (`entity-analysis-model-dictionary-kvp-test-bench`)

Hierarchy Placement: Level 4, child of Data Structures Dictionaries Supervisor
Semantic Persona: Owns the entity analysis model dictionary kvp slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: entity, analysis, model, dictionary, kvp, test, bench
Toolbox (Max 15):
- EntityAnalysisModelDictionaryKvpValidate: Validates a Dictionary KVP without saving it and returns every failure. (permissions from EntityAnalysisModelDictionaryKvpService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Dictionary Kvp Author (`entity-analysis-model-dictionary-kvp-author`)

Hierarchy Placement: Level 4, child of Data Structures Dictionaries Supervisor
Semantic Persona: Owns the entity analysis model dictionary kvp slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: entity, analysis, model, dictionary, kvp, author
Toolbox (Max 15):
- EntityAnalysisModelDictionaryKvpList: Lists Key Value Pairs for the caller’s tenant, keyset-paged and capped. ( permissions from EntityAnalysisModelDictionaryKvpService.RequiredPermissions)
- EntityAnalysisModelDictionaryKvpFilterFields: Lists the fields a query builder JSON filter over Dictionary KVPs may use. (permissions from EntityAnalysisModelDictionaryKvpService.RequiredPermissions)
- EntityAnalysisModelDictionaryKvpFilter: Returns the Dictionary KVPs matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelDictionaryKvpService.RequiredPermissions)
- EntityAnalysisModelDictionaryKvpCount: Counts the Dictionary KVPs matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelDictionaryKvpService.RequiredPermissions)
- EntityAnalysisModelDictionaryKvpGet: Returns one Key Value Pair, matched on the parent Dictionary’s id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelDictionaryKvpService.RequiredPermissions)
- EntityAnalysisModelDictionaryKvpGetByGuid: Returns one Key Value Pair, matched on the parent Dictionary’s id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelDictionaryKvpService.RequiredPermissions)
- EntityAnalysisModelDictionaryKvpGetByEntityAnalysisModelDictionaryId: Lists the Key Value Pairs belonging to a given Dictionary, ordered by Id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelDictionaryKvpService.RequiredPermissions)
- EntityAnalysisModelDictionaryKvpValidate: Validates a Dictionary KVP without saving it and returns every failure. (permissions from EntityAnalysisModelDictionaryKvpService.RequiredPermissions)
- EntityAnalysisModelDictionaryKvpCreate: Registers a Key Value Pair under a Dictionary in the caller’s tenant; calling twice creates two. Write — confirmation required (permissions from EntityAnalysisModelDictionaryKvpService.RequiredPermissions)
- EntityAnalysisModelDictionaryKvpUpdate: Updates a Key Value Pair in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelDictionaryKvpService.RequiredPermissions)
- EntityAnalysisModelDictionaryKvpDelete: Soft-deletes a Key Value Pair in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelDictionaryKvpService.RequiredPermissions)
- EntityAnalysisModelDictionaryKvpUploadCsv: Imports CSV files of key,value[,deleteExpiryDate] lines into a Dictionary in the caller’s tenant; existing keys are updated, lines that fail are skipped, and one upload record is written per file. Write — confirmation required (permissions from EntityAnalysisModelDictionaryKvpService.RequiredPermissions)
- EntityAnalysisModelDictionaryKvpResubmit: Resubmits a rejected Key Value Pair for approval again, unchanged, opening a fresh approval cycle. Write — confirmation required (permissions from EntityAnalysisModelDictionaryKvpService.RequiredPermissions)
- EntityAnalysisModelDictionaryKvpRetract: Retracts a rejected Key Value Pair, restoring it to its last approved version without a new approval cycle. Write — confirmation required (permissions from EntityAnalysisModelDictionaryKvpService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Data Structures Lists Supervisor

Hierarchy Placement: Level 3, child of Model Supporting Structures
Semantic Persona: Routes within the data structures lists slice of model supporting structures. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: data structures lists, entity, analysis, model, list, value
Toolbox (Max 15): none — routes to 6 leaf agents covering 21 tools
Downstream Dependencies: entity-analysis-model-list-inspector, entity-analysis-model-list-test-bench, entity-analysis-model-list-author, entity-analysis-model-list-value-inspector, entity-analysis-model-list-value-test-bench, entity-analysis-model-list-value-author

Entity Analysis Model List Inspector (`entity-analysis-model-list-inspector`)

Hierarchy Placement: Level 4, child of Data Structures Lists Supervisor
Semantic Persona: Owns the entity analysis model list slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: entity, analysis, model, list, inspector
Toolbox (Max 15):
- EntityAnalysisModelListList: Lists Lists for the caller’s tenant, keyset-paged and capped. (permissions from EntityAnalysisModelListService.RequiredPermissions)
- EntityAnalysisModelListFilterFields: Lists the fields a query builder JSON filter over Lists may use. ( permissions from EntityAnalysisModelListService.RequiredPermissions)
- EntityAnalysisModelListFilter: Returns the Lists matching query builder JSON, keyset-paged and capped. ( permissions from EntityAnalysisModelListService.RequiredPermissions)
- EntityAnalysisModelListCount: Counts the Lists matching query builder JSON, optionally grouped by a field. ( permissions from EntityAnalysisModelListService.RequiredPermissions)
- EntityAnalysisModelListGet: Returns one List by id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelListService.RequiredPermissions)
- EntityAnalysisModelListGetByGuid: Returns one List by Guid, scoped to the caller’s tenant. (permissions from EntityAnalysisModelListService.RequiredPermissions)
- EntityAnalysisModelListGetByEntityAnalysisModelId: Lists the Lists belonging to a given Model, ordered by Id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelListService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model List Test Bench (`entity-analysis-model-list-test-bench`)

Hierarchy Placement: Level 4, child of Data Structures Lists Supervisor
Semantic Persona: Owns the entity analysis model list slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: entity, analysis, model, list, test, bench
Toolbox (Max 15):
- EntityAnalysisModelListValidate: Validates a List without saving it and returns every failure. (permissions from EntityAnalysisModelListService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model List Author (`entity-analysis-model-list-author`)

Hierarchy Placement: Level 4, child of Data Structures Lists Supervisor
Semantic Persona: Owns the entity analysis model list slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: entity, analysis, model, list, author
Toolbox (Max 15):
- EntityAnalysisModelListList: Lists Lists for the caller’s tenant, keyset-paged and capped. (permissions from EntityAnalysisModelListService.RequiredPermissions)
- EntityAnalysisModelListFilterFields: Lists the fields a query builder JSON filter over Lists may use. ( permissions from EntityAnalysisModelListService.RequiredPermissions)
- EntityAnalysisModelListFilter: Returns the Lists matching query builder JSON, keyset-paged and capped. ( permissions from EntityAnalysisModelListService.RequiredPermissions)
- EntityAnalysisModelListCount: Counts the Lists matching query builder JSON, optionally grouped by a field. ( permissions from EntityAnalysisModelListService.RequiredPermissions)
- EntityAnalysisModelListGet: Returns one List by id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelListService.RequiredPermissions)
- EntityAnalysisModelListGetByGuid: Returns one List by Guid, scoped to the caller’s tenant. (permissions from EntityAnalysisModelListService.RequiredPermissions)
- EntityAnalysisModelListGetByEntityAnalysisModelId: Lists the Lists belonging to a given Model, ordered by Id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelListService.RequiredPermissions)
- EntityAnalysisModelListValidate: Validates a List without saving it and returns every failure. (permissions from EntityAnalysisModelListService.RequiredPermissions)
- EntityAnalysisModelListCreate: Registers a List under a Model in the caller’s tenant; calling twice creates two. Write — confirmation required (permissions from EntityAnalysisModelListService.RequiredPermissions)
- EntityAnalysisModelListUpdate: Updates a List in the caller’s tenant by id. Write — confirmation required ( permissions from EntityAnalysisModelListService.RequiredPermissions)
- EntityAnalysisModelListDelete: Soft-deletes a List in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelListService.RequiredPermissions)
- EntityAnalysisModelListResubmit: Resubmits a rejected List for approval again, unchanged, opening a fresh approval cycle. Write — confirmation required (permissions from EntityAnalysisModelListService.RequiredPermissions)
- EntityAnalysisModelListRetract: Retracts a rejected List, restoring it to its last approved version without a new approval cycle. Write — confirmation required (permissions from EntityAnalysisModelListService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model List Value Inspector (`entity-analysis-model-list-value-inspector`)

Hierarchy Placement: Level 4, child of Data Structures Lists Supervisor
Semantic Persona: Owns the entity analysis model list value slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: entity, analysis, model, list, value, inspector
Toolbox (Max 15):
- EntityAnalysisModelListValueList: Lists List Values for the caller’s tenant, keyset-paged and capped. ( permissions from EntityAnalysisModelListValueService.RequiredPermissions)
- EntityAnalysisModelListValueFilterFields: Lists the fields a query builder JSON filter over List Values may use. (permissions from EntityAnalysisModelListValueService.RequiredPermissions)
- EntityAnalysisModelListValueFilter: Returns the List Values matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelListValueService.RequiredPermissions)
- EntityAnalysisModelListValueCount: Counts the List Values matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelListValueService.RequiredPermissions)
- EntityAnalysisModelListValueGet: Returns one List Value matched against its parent List’s id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelListValueService.RequiredPermissions)
- EntityAnalysisModelListValueGetByGuid: Returns one List Value matched against its parent List’s id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelListValueService.RequiredPermissions)
- EntityAnalysisModelListValueGetByEntityAnalysisModelListId: Lists the values belonging to a given List, ordered by Id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelListValueService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model List Value Test Bench (`entity-analysis-model-list-value-test-bench`)

Hierarchy Placement: Level 4, child of Data Structures Lists Supervisor
Semantic Persona: Owns the entity analysis model list value slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: entity, analysis, model, list, value, test, bench
Toolbox (Max 15):
- EntityAnalysisModelListValueValidate: Validates a List Value without saving it and returns every failure. ( permissions from EntityAnalysisModelListValueService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model List Value Author (`entity-analysis-model-list-value-author`)

Hierarchy Placement: Level 4, child of Data Structures Lists Supervisor
Semantic Persona: Owns the entity analysis model list value slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: entity, analysis, model, list, value, author
Toolbox (Max 15):
- EntityAnalysisModelListValueList: Lists List Values for the caller’s tenant, keyset-paged and capped. ( permissions from EntityAnalysisModelListValueService.RequiredPermissions)
- EntityAnalysisModelListValueFilterFields: Lists the fields a query builder JSON filter over List Values may use. (permissions from EntityAnalysisModelListValueService.RequiredPermissions)
- EntityAnalysisModelListValueFilter: Returns the List Values matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelListValueService.RequiredPermissions)
- EntityAnalysisModelListValueCount: Counts the List Values matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelListValueService.RequiredPermissions)
- EntityAnalysisModelListValueGet: Returns one List Value matched against its parent List’s id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelListValueService.RequiredPermissions)
- EntityAnalysisModelListValueGetByGuid: Returns one List Value matched against its parent List’s id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelListValueService.RequiredPermissions)
- EntityAnalysisModelListValueGetByEntityAnalysisModelListId: Lists the values belonging to a given List, ordered by Id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelListValueService.RequiredPermissions)
- EntityAnalysisModelListValueValidate: Validates a List Value without saving it and returns every failure. ( permissions from EntityAnalysisModelListValueService.RequiredPermissions)
- EntityAnalysisModelListValueCreate: Registers a value under a List in the caller’s tenant; calling twice creates two. Write — confirmation required (permissions from EntityAnalysisModelListValueService.RequiredPermissions)
- EntityAnalysisModelListValueUpdate: Updates a value in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelListValueService.RequiredPermissions)
- EntityAnalysisModelListValueDelete: Soft-deletes a value in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelListValueService.RequiredPermissions)
- EntityAnalysisModelListValueUploadCsv: Loads one or more CSV files of values into a List in the caller’s tenant in a single transaction, recording an upload entry per file; calling twice loads the values twice. Write — confirmation required (permissions from EntityAnalysisModelListValueService.RequiredPermissions)
- EntityAnalysisModelListValueResubmit: Resubmits a rejected List Value for approval again, unchanged, opening a fresh approval cycle. Write — confirmation required (permissions from EntityAnalysisModelListValueService.RequiredPermissions)
- EntityAnalysisModelListValueRetract: Retracts a rejected List Value, restoring it to its last approved version without a new approval cycle. Write — confirmation required (permissions from EntityAnalysisModelListValueService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Hierarchy Placement: Level 3, child of Model Supporting Structures
Semantic Persona: Routes within the navigation slice of model supporting structures. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: navigation, tree, children
Toolbox (Max 15): none — routes to 4 leaf agents covering 27 tools
Downstream Dependencies: tree-children-inspector-1, tree-children-inspector-2, tree-children-inspector-3, tree-children-inspector-4

Tree Children Inspector 1 (`tree-children-inspector-1`)

Hierarchy Placement: Level 4, child of Navigation Supervisor
Semantic Persona: Owns the tree children slice end to end, which is small enough that reads and writes stay with one agent.
Semantic Routing Keywords: tree, children, inspector, 1
Toolbox (Max 15):
- TreeChildrenRequestXPathGet: Lists the request XPath nodes of an Entity Analysis Model as tree children, ordered by Id. (permissions from TreeChildrenNodes.RequestXPath)
- TreeChildrenVisualisationRegistryDatasourceGet: Lists the datasource nodes of a Visualisation Registry as tree children, ordered by priority. (permissions from TreeChildrenNodes.VisualisationRegistryDatasource)
- TreeChildrenUserRegistryGet: Lists the users of a Role Registry as tree children. (permissions from TreeChildrenNodes.UserRegistry)
- TreeChildrenRoleRegistryPermissionGet: Lists the permission nodes granted to a Role Registry as tree children; the name is the permission specification name. (permissions from TreeChildrenNodes.RoleRegistryPermission)
- TreeChildrenVisualisationRegistryParameterGet: Lists the parameter nodes of a Visualisation Registry as tree children, ordered by Id. (permissions from TreeChildrenNodes.VisualisationRegistryParameter)
- TreeChildrenInlineFunctionGet: Lists the inline function nodes of an Entity Analysis Model as tree children, ordered by Id. (permissions from TreeChildrenNodes.InlineFunction)
- TreeChildrenTagGet: Lists the tag nodes of an Entity Analysis Model as tree children, ordered by Id. ( permissions from TreeChildrenNodes.Tag)
Downstream Dependencies: none (leaf)

Tree Children Inspector 2 (`tree-children-inspector-2`)

Hierarchy Placement: Level 4, child of Navigation Supervisor
Semantic Persona: Owns the tree children slice end to end, which is small enough that reads and writes stay with one agent.
Semantic Routing Keywords: tree, children, inspector, 2
Toolbox (Max 15):
- TreeChildrenGatewayRuleGet: Lists the gateway rule nodes of an Entity Analysis Model as tree children, ordered by priority. (permissions from TreeChildrenNodes.GatewayRule)
- TreeChildrenExhaustiveGet: Lists the exhaustive search instance nodes of an Entity Analysis Model as tree children, ordered by Id. (permissions from TreeChildrenNodes.Exhaustive)
- TreeChildrenReprocessingGet: Lists the reprocessing rule nodes of an Entity Analysis Model as tree children. ( permissions from TreeChildrenNodes.Reprocessing)
- TreeChildrenAdaptationGet: Lists the HTTP adaptation nodes of an Entity Analysis Model as tree children, ordered by Id. (permissions from TreeChildrenNodes.Adaptation)
- TreeChildrenCaseWorkflowGet: Lists the case workflow nodes of an Entity Analysis Model as tree children, ordered by Id. (permissions from TreeChildrenNodes.CaseWorkflow)
- TreeChildrenAbstractionCalculationGet: Lists the abstraction calculation nodes of an Entity Analysis Model as tree children. (permissions from TreeChildrenNodes.AbstractionCalculation)
- TreeChildrenAbstractionRuleGet: Lists the abstraction rule nodes of an Entity Analysis Model as tree children. (permissions from TreeChildrenNodes.AbstractionRule)
Downstream Dependencies: none (leaf)

Tree Children Inspector 3 (`tree-children-inspector-3`)

Hierarchy Placement: Level 4, child of Navigation Supervisor
Semantic Persona: Owns the tree children slice end to end, which is small enough that reads and writes stay with one agent.
Semantic Routing Keywords: tree, children, inspector, 3
Toolbox (Max 15):
- TreeChildrenActivationRuleGet: Lists the activation rule nodes of an Entity Analysis Model as tree children, ordered by priority. (permissions from TreeChildrenNodes.ActivationRule)
- TreeChildrenTTLCounterGet: Lists the TTL counter nodes of an Entity Analysis Model as tree children, ordered by Id. (permissions from TreeChildrenNodes.TtlCounter)
- TreeChildrenInlineScriptGet: Lists the inline script nodes of an Entity Analysis Model as tree children, ordered by Id. (permissions from TreeChildrenNodes.InlineScript)
- TreeChildrenSanctionsGet: Lists the sanction nodes of an Entity Analysis Model as tree children, ordered by Id. (permissions from TreeChildrenNodes.Sanctions)
- TreeChildrenListGet: Lists the list nodes of an Entity Analysis Model, addressed by model Guid, as tree children. (permissions from TreeChildrenNodes.List)
- TreeChildrenDictionaryGet: Lists the dictionary nodes of an Entity Analysis Model, addressed by model Guid, as tree children. (permissions from TreeChildrenNodes.Dictionary)
- TreeChildrenCaseWorkflowXPathGet: Lists the xpath nodes of a Case Workflow as tree children. (permissions from TreeChildrenNodes.CaseWorkflowXPath)
Downstream Dependencies: none (leaf)

Tree Children Inspector 4 (`tree-children-inspector-4`)

Hierarchy Placement: Level 4, child of Navigation Supervisor
Semantic Persona: Owns the tree children slice end to end, which is small enough that reads and writes stay with one agent.
Semantic Routing Keywords: tree, children, inspector, 4
Toolbox (Max 15):
- TreeChildrenCaseWorkflowFormGet: Lists the form nodes of a Case Workflow as tree children. (permissions from TreeChildrenNodes.CaseWorkflowForm)
- TreeChildrenCaseWorkflowActionGet: Lists the action nodes of a Case Workflow as tree children. (permissions from TreeChildrenNodes.CaseWorkflowAction)
- TreeChildrenCaseWorkflowMacroGet: Lists the macro nodes of a Case Workflow as tree children. (permissions from TreeChildrenNodes.CaseWorkflowMacro)
- TreeChildrenCaseWorkflowFilterGet: Lists the filter nodes of a Case Workflow as tree children. (permissions from TreeChildrenNodes.CaseWorkflowFilter)
- TreeChildrenCaseWorkflowDisplayGet: Lists the display nodes of a Case Workflow as tree children. (permissions from TreeChildrenNodes.CaseWorkflowDisplay)
- TreeChildrenCaseWorkflowStatusGet: Lists the status nodes of a Case Workflow as tree children. (permissions from TreeChildrenNodes.CaseWorkflowStatus)
Downstream Dependencies: none (leaf)

Overrides Supervisor

Hierarchy Placement: Level 3, child of Model Supporting Structures
Semantic Persona: Routes within the overrides slice of model supporting structures. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: overrides, entity, analysis, model, override, query
Toolbox (Max 15): none — routes to 4 leaf agents covering 14 tools
Downstream Dependencies: entity-analysis-model-override-inspector, entity-analysis-model-override-test-bench, entity-analysis-model-override-author, entity-analysis-model-override-query

Entity Analysis Model Override Inspector (`entity-analysis-model-override-inspector`)

Hierarchy Placement: Level 4, child of Overrides Supervisor
Semantic Persona: Owns the entity analysis model override slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: entity, analysis, model, override, inspector
Toolbox (Max 15):
- EntityAnalysisModelOverrideList: Lists Overrides for the caller’s tenant, keyset-paged and capped. (permissions from EntityAnalysisModelOverrideService.RequiredPermissions)
- EntityAnalysisModelOverrideFilterFields: Lists the fields a query builder JSON filter over Overrides may use. ( permissions from EntityAnalysisModelOverrideService.RequiredPermissions)
- EntityAnalysisModelOverrideFilter: Returns the Overrides matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelOverrideService.RequiredPermissions)
- EntityAnalysisModelOverrideCount: Counts the Overrides matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelOverrideService.RequiredPermissions)
- EntityAnalysisModelOverrideGet: Returns one Override by id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelOverrideService.RequiredPermissions)
- EntityAnalysisModelOverrideGetByEntityAnalysisModelId: Lists Overrides belonging to a given Model, scoped to the caller’s tenant. (permissions from EntityAnalysisModelOverrideService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Override Test Bench (`entity-analysis-model-override-test-bench`)

Hierarchy Placement: Level 4, child of Overrides Supervisor
Semantic Persona: Owns the entity analysis model override slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: entity, analysis, model, override, test, bench
Toolbox (Max 15):
- EntityAnalysisModelOverrideValidate: Validates a Override without saving it and returns every failure. ( permissions from EntityAnalysisModelOverrideService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Override Author (`entity-analysis-model-override-author`)

Hierarchy Placement: Level 4, child of Overrides Supervisor
Semantic Persona: Owns the entity analysis model override slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: entity, analysis, model, override, author
Toolbox (Max 15):
- EntityAnalysisModelOverrideList: Lists Overrides for the caller’s tenant, keyset-paged and capped. (permissions from EntityAnalysisModelOverrideService.RequiredPermissions)
- EntityAnalysisModelOverrideFilterFields: Lists the fields a query builder JSON filter over Overrides may use. ( permissions from EntityAnalysisModelOverrideService.RequiredPermissions)
- EntityAnalysisModelOverrideFilter: Returns the Overrides matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelOverrideService.RequiredPermissions)
- EntityAnalysisModelOverrideCount: Counts the Overrides matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelOverrideService.RequiredPermissions)
- EntityAnalysisModelOverrideGet: Returns one Override by id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelOverrideService.RequiredPermissions)
- EntityAnalysisModelOverrideGetByEntityAnalysisModelId: Lists Overrides belonging to a given Model, scoped to the caller’s tenant. (permissions from EntityAnalysisModelOverrideService.RequiredPermissions)
- EntityAnalysisModelOverrideValidate: Validates a Override without saving it and returns every failure. ( permissions from EntityAnalysisModelOverrideService.RequiredPermissions)
- EntityAnalysisModelOverrideCreate: Registers a Override against a Model in the caller’s tenant; calling twice creates two. Write — confirmation required (permissions from EntityAnalysisModelOverrideService.RequiredPermissions)
- EntityAnalysisModelOverrideUpdate: Toggles a Override on or off by row existence; NOT idempotent – repeat calls flip state. Write — confirmation required (permissions from EntityAnalysisModelOverrideService.RequiredPermissions)
- EntityAnalysisModelOverrideUpdateDeleteExpiryDate: Updates the Delete Expiry Date of an existing active Override in the caller’s tenant. Write — confirmation required (permissions from EntityAnalysisModelOverrideService.RequiredPermissions)
- EntityAnalysisModelOverrideDelete: Soft-deletes a Override in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelOverrideService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Override Query (`entity-analysis-model-override-query`)

Hierarchy Placement: Level 4, child of Overrides Supervisor
Semantic Persona: Owns the entity analysis model override query slice end to end, which is small enough that reads and writes stay with one agent.
Semantic Routing Keywords: entity, analysis, model, override, query
Toolbox (Max 15):
- EntityAnalysisModelOverrideQueryGet: Lists tenant models supporting override on a key, flagging active overrides for a key value. (permissions from EntityAnalysisModelOverrideQueryService.RequiredPermissions)
- EntityAnalysisModelOverrideQueryGetKeys: Lists the override keys available for a Model in the caller’s tenant. (permissions from EntityAnalysisModelOverrideQueryService.RequiredPermissions)
- EntityAnalysisModelOverrideQueryGetValues: Lists the values held against an override key for a Model in the caller’s tenant. (permissions from EntityAnalysisModelOverrideQueryService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Reprocessing Supervisor

Hierarchy Placement: Level 3, child of Model Supporting Structures
Semantic Persona: Routes within the reprocessing slice of model supporting structures. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: reprocessing, entity, analysis, model, rule, instance
Toolbox (Max 15): none — routes to 6 leaf agents covering 24 tools
Downstream Dependencies: entity-analysis-model-reprocessing-rule-inspector, entity-analysis-model-reprocessing-rule-test-bench, entity-analysis-model-reprocessing-rule-author, entity-analysis-model-reprocessing-rule-instance-inspector, entity-analysis-model-reprocessing-rule-instance-test-bench, entity-analysis-model-reprocessing-rule-instance-author
Cross-Domain Reach: completions, model-invocation-context-and-dependencies

Entity Analysis Model Reprocessing Rule Inspector (`entity-analysis-model-reprocessing-rule-inspector`)

Hierarchy Placement: Level 4, child of Reprocessing Supervisor
Semantic Persona: Owns the entity analysis model reprocessing rule slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: entity, analysis, model, reprocessing, rule, inspector
Toolbox (Max 15):
- EntityAnalysisModelReprocessingRuleList: Lists Reprocessing Rules for the caller’s tenant, keyset-paged and capped. (permissions from EntityAnalysisModelReprocessingRuleService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleFilterFields: Lists the fields a query builder JSON filter over Reprocessing Rules may use. (permissions from EntityAnalysisModelReprocessingRuleService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleFilter: Returns the Reprocessing Rules matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelReprocessingRuleService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleCount: Counts the Reprocessing Rules matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelReprocessingRuleService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleGet: Returns one Reprocessing Rule by id, scoped to the caller’s tenant. ( permissions from EntityAnalysisModelReprocessingRuleService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleGetByEntityAnalysisModelId: Lists Reprocessing Rules belonging to a given Model, scoped to the caller’s tenant. (permissions from EntityAnalysisModelReprocessingRuleService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Reprocessing Rule Test Bench (`entity-analysis-model-reprocessing-rule-test-bench`)

Hierarchy Placement: Level 4, child of Reprocessing Supervisor
Semantic Persona: Owns the entity analysis model reprocessing rule slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: entity, analysis, model, reprocessing, rule, test, bench
Toolbox (Max 15):
- EntityAnalysisModelReprocessingRuleValidate: Validates a Reprocessing Rule without saving it and returns every failure. (permissions from EntityAnalysisModelReprocessingRuleService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleParseRule: Parses and compiles a Reprocessing Rule’s rule text as the engine would and returns each located error. (permissions from EntityAnalysisModelReprocessingRuleService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleExecute: Runs a Reprocessing Rule against an invocation context as the engine would and returns the result. (permissions from EntityAnalysisModelReprocessingRuleService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleBuildRuleFromBuilderJson: Turns query builder JSON into a Reprocessing Rule’s rule text as the browser builder would, and parses it. (permissions from EntityAnalysisModelReprocessingRuleService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Reprocessing Rule Author (`entity-analysis-model-reprocessing-rule-author`)

Hierarchy Placement: Level 4, child of Reprocessing Supervisor
Semantic Persona: Owns the entity analysis model reprocessing rule slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: entity, analysis, model, reprocessing, rule, author
Toolbox (Max 15):
- EntityAnalysisModelReprocessingRuleList: Lists Reprocessing Rules for the caller’s tenant, keyset-paged and capped. (permissions from EntityAnalysisModelReprocessingRuleService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleFilterFields: Lists the fields a query builder JSON filter over Reprocessing Rules may use. (permissions from EntityAnalysisModelReprocessingRuleService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleFilter: Returns the Reprocessing Rules matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelReprocessingRuleService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleCount: Counts the Reprocessing Rules matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelReprocessingRuleService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleGet: Returns one Reprocessing Rule by id, scoped to the caller’s tenant. ( permissions from EntityAnalysisModelReprocessingRuleService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleGetByEntityAnalysisModelId: Lists Reprocessing Rules belonging to a given Model, scoped to the caller’s tenant. (permissions from EntityAnalysisModelReprocessingRuleService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleValidate: Validates a Reprocessing Rule without saving it and returns every failure. (permissions from EntityAnalysisModelReprocessingRuleService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleParseRule: Parses and compiles a Reprocessing Rule’s rule text as the engine would and returns each located error. (permissions from EntityAnalysisModelReprocessingRuleService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleExecute: Runs a Reprocessing Rule against an invocation context as the engine would and returns the result. (permissions from EntityAnalysisModelReprocessingRuleService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleBuildRuleFromBuilderJson: Turns query builder JSON into a Reprocessing Rule’s rule text as the browser builder would, and parses it. (permissions from EntityAnalysisModelReprocessingRuleService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleCreate: Registers a Reprocessing Rule under a Model in the caller’s tenant; calling twice creates two. Write — confirmation required (permissions from EntityAnalysisModelReprocessingRuleService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleUpdate: Supersedes a Reprocessing Rule in the caller’s tenant by id, assigning it a new id. Write — confirmation required (permissions from EntityAnalysisModelReprocessingRuleService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleDelete: Soft-deletes a Reprocessing Rule in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelReprocessingRuleService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Reprocessing Rule Instance Inspector (`entity-analysis-model-reprocessing-rule-instance-inspector`)

Hierarchy Placement: Level 4, child of Reprocessing Supervisor
Semantic Persona: Owns the entity analysis model reprocessing rule instance slice and reads only; it never changes anything and is the safe default for any question about current state.
Semantic Routing Keywords: entity, analysis, model, reprocessing, rule, instance, inspector
Toolbox (Max 15):
- EntityAnalysisModelReprocessingRuleInstanceList: Lists Reprocessing Rule instances for the caller’s tenant, keyset-paged and capped. (permissions from EntityAnalysisModelReprocessingRuleInstanceService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleInstanceFilterFields: Lists the fields a query builder JSON filter over Reprocessing Rule Instances may use. (permissions from EntityAnalysisModelReprocessingRuleInstanceService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleInstanceFilter: Returns the Reprocessing Rule Instances matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelReprocessingRuleInstanceService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleInstanceCount: Counts the Reprocessing Rule Instances matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelReprocessingRuleInstanceService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleInstanceGet: Returns one Reprocessing Rule instance by id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelReprocessingRuleInstanceService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleInstanceGetByEntityAnalysisModelReprocessingId: Lists Reprocessing Rule instances belonging to a given Reprocessing Rule, scoped to the caller’s tenant. (permissions from EntityAnalysisModelReprocessingRuleInstanceService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Reprocessing Rule Instance Test Bench (`entity-analysis-model-reprocessing-rule-instance-test-bench`)

Hierarchy Placement: Level 4, child of Reprocessing Supervisor
Semantic Persona: Owns the entity analysis model reprocessing rule instance slice and runs the side-effect-free rehearsals: validation, rule parsing and trial execution, so a change can be proven before it is made.
Semantic Routing Keywords: entity, analysis, model, reprocessing, rule, instance, test, bench
Toolbox (Max 15):
- EntityAnalysisModelReprocessingRuleInstanceValidate: Validates a Reprocessing Rule Instance without saving it and returns every failure. (permissions from EntityAnalysisModelReprocessingRuleInstanceService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Entity Analysis Model Reprocessing Rule Instance Author (`entity-analysis-model-reprocessing-rule-instance-author`)

Hierarchy Placement: Level 4, child of Reprocessing Supervisor
Semantic Persona: Owns the entity analysis model reprocessing rule instance slice and is the only agent that mutates this slice; every call needs explicit confirmation and, where the slice is preservation-managed, a pre-change snapshot.
Semantic Routing Keywords: entity, analysis, model, reprocessing, rule, instance, author
Toolbox (Max 15):
- EntityAnalysisModelReprocessingRuleInstanceList: Lists Reprocessing Rule instances for the caller’s tenant, keyset-paged and capped. (permissions from EntityAnalysisModelReprocessingRuleInstanceService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleInstanceFilterFields: Lists the fields a query builder JSON filter over Reprocessing Rule Instances may use. (permissions from EntityAnalysisModelReprocessingRuleInstanceService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleInstanceFilter: Returns the Reprocessing Rule Instances matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelReprocessingRuleInstanceService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleInstanceCount: Counts the Reprocessing Rule Instances matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelReprocessingRuleInstanceService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleInstanceGet: Returns one Reprocessing Rule instance by id, scoped to the caller’s tenant. (permissions from EntityAnalysisModelReprocessingRuleInstanceService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleInstanceGetByEntityAnalysisModelReprocessingId: Lists Reprocessing Rule instances belonging to a given Reprocessing Rule, scoped to the caller’s tenant. (permissions from EntityAnalysisModelReprocessingRuleInstanceService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleInstanceValidate: Validates a Reprocessing Rule Instance without saving it and returns every failure. (permissions from EntityAnalysisModelReprocessingRuleInstanceService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleInstanceCreate: Registers a Reprocessing Rule instance under a Reprocessing Rule in the caller’s tenant; calling twice creates two. Write — confirmation required (permissions from EntityAnalysisModelReprocessingRuleInstanceService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleInstanceInsertByExistingUpdateUncompleted: Triggers a new reprocessing run for a Reprocessing Rule; fails if one is already uncompleted. Write — confirmation required (permissions from EntityAnalysisModelReprocessingRuleInstanceService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleInstanceUpdate: Supersedes a Reprocessing Rule instance in the caller’s tenant by id, assigning it a new id. Write — confirmation required (permissions from EntityAnalysisModelReprocessingRuleInstanceService.RequiredPermissions)
- EntityAnalysisModelReprocessingRuleInstanceDelete: Soft-deletes a Reprocessing Rule instance in the caller’s tenant by id. Write — confirmation required (permissions from EntityAnalysisModelReprocessingRuleInstanceService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Platform Reliability (`platform-ops`)

Platform Reliability

Hierarchy Placement: Level 2, child of Jooby
Semantic Persona: Reads the operational health of the platform: Postgres, Redis, etcd and Patroni, Docker and the overlay network, OpenTelemetry, logs and engine throughput counters. Almost entirely landlord-only and almost entirely read-only.
Semantic Routing Keywords: postgres, redis, etcd, patroni, docker, haproxy, overlay network, telemetry, OTLP, log, queue balance, latency, throughput, replication lag, WAF
Toolbox (Max 15): none — routes to 2 supervisors and 5 leaf agents covering 59 tools
Downstream Dependencies: platform-ops.engine-health, platform-ops.telemetry, application-and-container-logs, containers-load-balancer-and-overlay-network, etcd-and-patroni-cluster-state, postgres-observability, redis-observability

Engine Health Supervisor

Hierarchy Placement: Level 3, child of Platform Reliability
Semantic Persona: Routes within the engine health slice of platform reliability. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: engine health, entity, analysis, model, synchronisation, schedule, capture, queue, health, asynchronous, balance, processing, counter, response, time, pipeline, stage, performance, node, status, entries, task, http, invoke, warning
Toolbox (Max 15): none — routes to 3 leaf agents covering 17 tools
Downstream Dependencies: entity-analysis-model-synchronisation-schedule, engine-queue-and-throughput-counters, invoke-path-performance-and-warnings

Entity Analysis Model Synchronisation Schedule (`entity-analysis-model-synchronisation-schedule`)

Hierarchy Placement: Level 4, child of Engine Health Supervisor
Semantic Persona: Owns the entity analysis model synchronisation schedule slice end to end, which is small enough that reads and writes stay with one agent. It holds 1 mutating tool, so a confirmation is required before any of them runs.
Semantic Routing Keywords: entity, analysis, model, synchronisation, schedule
Toolbox (Max 15):
- EntityAnalysisModelSynchronisationScheduleGetCurrent: Returns the most recently scheduled model-synchronisation row for the caller’s tenant. (permissions from EntityAnalysisModelSynchronisationScheduleService.RequiredPermissions)
- EntityAnalysisModelSynchronisationScheduleList: Lists scheduled model-synchronisation rows for the caller’s tenant, keyset-paged and capped. (permissions from EntityAnalysisModelSynchronisationScheduleService.RequiredPermissions)
- EntityAnalysisModelSynchronisationScheduleFilterFields: Lists the fields a query builder JSON filter over Synchronisation Schedules may use. (permissions from EntityAnalysisModelSynchronisationScheduleService.RequiredPermissions)
- EntityAnalysisModelSynchronisationScheduleFilter: Returns the Synchronisation Schedules matching query builder JSON, keyset-paged and capped. (permissions from EntityAnalysisModelSynchronisationScheduleService.RequiredPermissions)
- EntityAnalysisModelSynchronisationScheduleCount: Counts the Synchronisation Schedules matching query builder JSON, optionally grouped by a field. (permissions from EntityAnalysisModelSynchronisationScheduleService.RequiredPermissions)
- EntityAnalysisModelSynchronisationScheduleCreate: Schedules model synchronisation across the cluster; calling twice creates two rows. Write — confirmation required (permissions from EntityAnalysisModelSynchronisationScheduleService.RequiredPermissions)
- EntityAnalysisModelSynchronisationScheduleValidate: Validates a Synchronisation Schedule without saving it and returns every failure. (permissions from EntityAnalysisModelSynchronisationScheduleService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Engine Queue and Throughput Counters (`engine-queue-and-throughput-counters`)

Hierarchy Placement: Level 4, child of Engine Health Supervisor
Semantic Persona: Owns the engine queue and throughput counters slice end to end, which is small enough that reads and writes stay with one agent.
Semantic Routing Keywords: capture, queue, health, entity, analysis, asynchronous, balance, model, processing, counter, response, time, pipeline, stage, performance, synchronisation, node, status, entries
Toolbox (Max 15):
- CaptureQueueHealthList: Lists per-cycle depth and dropped-count snapshots for every bounded in-memory capture queue in the observability suite (ModelInvokeWarning, CaseCreationWarning, ArchiverWarning, RedisSentinelEvent, RedisConnectionEvent, OpenTelemetryMetric)… (permissions from CaptureQueueHealthService.RequiredPermissions)
- EntityAnalysisAsynchronousQueueBalanceList: Lists platform-wide asynchronous processing queue balance rows, most recent first, capped at 100000, with optional date-range filters. (permissions from EntityAnalysisAsynchronousQueueBalanceService.RequiredPermissions)
- EntityAnalysisModelAsynchronousQueueBalanceList: Lists per-Model asynchronous processing queue balance rows, most recent first, capped at 100000, with optional date-range and Model-name substring filters. (permissions from EntityAnalysisModelAsynchronousQueueBalanceService.RequiredPermissions)
- EntityAnalysisModelProcessingCounterList: Lists per-Model processing counter rows, most recent first, capped at 100000, with optional date-range and Model-name substring filters. (permissions from EntityAnalysisModelProcessingCounterService.RequiredPermissions)
- EntityAnalysisModelResponseTimePipelineCounterList: Lists per-minute Response Time Pipeline Counter rows for a Model, most recent first, capped at 100000, with optional date-range and substring search (StageName) filters, scoped to the caller’s tenant. (permissions from EntityAnalysisModelResponseTimePipelineCounterService.RequiredPermissions)
- EntityAnalysisModelStagePerformanceCounterList: Lists per-minute Stage Performance Counter rows for a Model, most recent first, capped at 100000, with optional date-range and substring search (StageName) filters, scoped to the caller’s tenant. (permissions from EntityAnalysisModelStagePerformanceCounterService.RequiredPermissions)
- EntityAnalysisModelSynchronisationNodeStatusEntriesGet: Lists the tenant’s model-synchronisation node statuses seen in the last hour, with pending and availability flags. (permissions from EntityAnalysisModelSynchronisationNodeStatusEntriesService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Invoke Path Performance and Warnings (`invoke-path-performance-and-warnings`)

Hierarchy Placement: Level 4, child of Engine Health Supervisor
Semantic Persona: Owns the invoke path performance and warnings slice end to end, which is small enough that reads and writes stay with one agent.
Semantic Routing Keywords: entity, analysis, model, task, performance, counter, http, processing, invoke, warning
Toolbox (Max 15):
- EntityAnalysisModelTaskPerformanceCounterList: Lists per-minute Task Performance Counter rows for a Model, most recent first, capped at 100000, with optional date-range and substring search (TaskName/Direction) filters, scoped to the caller’s tenant – useful for cloud compute-cost attr… (permissions from EntityAnalysisModelTaskPerformanceCounterService.RequiredPermissions)
- HttpProcessingCounterList: Lists platform-wide HTTP processing counters, most recent first, capped at 100000, with optional date-range filters. (permissions from HttpProcessingCounterService.RequiredPermissions)
- ModelInvokeWarningList: Lists captured invocation warn-threshold breaches – the trace point message and timing gap that exceeded a model’s Warn Threshold Milliseconds, most recent first, capped at 100000, with optional model/date-range/substring filters. (permissions from ModelInvokeWarningService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Telemetry Supervisor

Hierarchy Placement: Level 3, child of Platform Reliability
Semantic Persona: Routes within the telemetry slice of platform reliability. Holds no tools; it exists so the domain above it chooses between a handful of coherent specialists rather than dozens of leaves.
Semantic Routing Keywords: telemetry, open, exclude, log, counter, dot, net, runtime, metric, otlp, dispatch
Toolbox (Max 15): none — routes to 3 leaf agents covering 15 tools
Downstream Dependencies: open-telemetry-exclude, telemetry-log-counters-and-runtime-metrics, telemetry-metrics-and-otlp-dispatch

Open Telemetry Exclude (`open-telemetry-exclude`)

Hierarchy Placement: Level 4, child of Telemetry Supervisor
Semantic Persona: Owns the open telemetry exclude slice end to end, which is small enough that reads and writes stay with one agent. It holds 3 mutating tools, so a confirmation is required before any of them runs.
Semantic Routing Keywords: open, telemetry, exclude
Toolbox (Max 15):
- OpenTelemetryExcludeList: Lists every OpenTelemetryExclude row, most recent first. Not scoped to any tenant. ( permissions from OpenTelemetryExcludeService.RequiredPermissions)
- OpenTelemetryExcludeGet: Returns one OpenTelemetryExclude row by id. (permissions from OpenTelemetryExcludeService.RequiredPermissions)
- OpenTelemetryExcludeCreate: Registers an OpenTelemetry instrument/counter name to stop exporting; calling twice creates two rows. Write — confirmation required (permissions from OpenTelemetryExcludeService.RequiredPermissions)
- OpenTelemetryExcludeValidate: Validates an Open Telemetry Exclude without saving it and returns every failure. (permissions from OpenTelemetryExcludeService.RequiredPermissions)
- OpenTelemetryExcludeUpdate: Updates an existing exclusion row in place. Write — confirmation required ( permissions from OpenTelemetryExcludeService.RequiredPermissions)
- OpenTelemetryExcludeDelete: Soft-deletes an exclusion row by id; the named instrument becomes eligible for export again. Write — confirmation required (permissions from OpenTelemetryExcludeService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Telemetry Log Counters and Runtime Metrics (`telemetry-log-counters-and-runtime-metrics`)

Hierarchy Placement: Level 4, child of Telemetry Supervisor
Semantic Persona: Owns the telemetry log counters and runtime metrics slice end to end, which is small enough that reads and writes stay with one agent. It holds 3 mutating tools, so a confirmation is required before any of them runs.
Semantic Routing Keywords: open, telemetry, log, counter, dot, net, runtime, metric
Toolbox (Max 15):
- OpenTelemetryLogCounterList: Lists every OpenTelemetryLogCounter rule (Name + Regex), most recent first. Not scoped to any tenant. (permissions from OpenTelemetryLogCounterService.RequiredPermissions)
- OpenTelemetryLogCounterGet: Returns one OpenTelemetryLogCounter rule by id. (permissions from OpenTelemetryLogCounterService.RequiredPermissions)
- OpenTelemetryLogCounterCreate: Registers a rule that increments an OpenTelemetry counter named Name whenever Regex matches an incoming unstructured log line; calling twice creates two rows. Write — confirmation required (permissions from OpenTelemetryLogCounterService.RequiredPermissions)
- OpenTelemetryLogCounterValidate: Validates an Open Telemetry Log Counter without saving it and returns every failure. (permissions from OpenTelemetryLogCounterService.RequiredPermissions)
- OpenTelemetryLogCounterUpdate: Updates an existing rule in place. Write — confirmation required ( permissions from OpenTelemetryLogCounterService.RequiredPermissions)
- OpenTelemetryLogCounterDelete: Soft-deletes a rule by id; it immediately stops being checked against incoming log lines. Write — confirmation required (permissions from OpenTelemetryLogCounterService.RequiredPermissions)
- DotNetRuntimeMetricList: Lists per-minute .NET runtime metric snapshots across every node, most recent first, capped at 100000, with optional date-range and substring search (Instance) filters, not scoped to any tenant. ( permissions from DotNetRuntimeMetricService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Telemetry Metrics and OTLP Dispatch (`telemetry-metrics-and-otlp-dispatch`)

Hierarchy Placement: Level 4, child of Telemetry Supervisor
Semantic Persona: Owns the telemetry metrics and otlp dispatch slice end to end, which is small enough that reads and writes stay with one agent.
Semantic Routing Keywords: open, telemetry, metric, otlp, dispatch, counter
Toolbox (Max 15):
- OpenTelemetryMetricList: Lists locally-aggregated OpenTelemetry metrics – everything already recorded onto ServiceDiagnostics/EngineDiagnostics’ Meters (jube.service.operation. , jube.engine.stage.duration, jube.engine.sampler., …), summarised once a minute int… (permissions from OpenTelemetryMetricService.RequiredPermissions)
- OtlpDispatchCounterList: Lists per-minute OTLP export dispatch counters – one row per distinct Signal ( “traces”/”metrics”/”logs”) per minute, with Count/SuccessCount/FailureCount/ItemCount/DroppedCount and Total/Min/MaxMicroseconds for that window, most recent fir… (permissions from OtlpDispatchCounterService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Application and Container Logs (`application-and-container-logs`)

Hierarchy Placement: Level 3, child of Platform Reliability
Semantic Persona: Owns the application and container logs slice end to end, which is small enough that reads and writes stay with one agent.
Semantic Routing Keywords: application, log, entry, container, waf, attack
Toolbox (Max 15):
- ApplicationLogEntryList: Lists application-wide WARN/ERROR/FATAL log entries, most recent first, capped at 100000, with optional date-range and substring search (Message/LoggerName/Exception) filters. (permissions from ApplicationLogEntryService.RequiredPermissions)
- ContainerLogEntryList: Lists every container’s own stdout/stderr on the Docker host (no name filtering – covers etcd/Patroni/Redis/Sentinel and every other container), most recent first, capped at 100000, with optional date-range and substring search (ContainerN… (permissions from ContainerLogEntryService.RequiredPermissions)
- WafAttackList: Lists captured Web Application Firewall signature matches, most recent first, capped at 100000, with optional date-range and substring search (MatchedValue/Route/SignatureName/Category/MatchedField/UserName/RemoteIp) filters. (permissions from WafAttackService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Containers, Load Balancer and Overlay Network (`containers-load-balancer-and-overlay-network`)

Hierarchy Placement: Level 3, child of Platform Reliability
Semantic Persona: Owns the containers, load balancer and overlay network slice end to end, which is small enough that reads and writes stay with one agent.
Semantic Routing Keywords: docker, container, metric, event, host, ha, proxy, reachability, probe, server, status, overlay, network, task, drift
Toolbox (Max 15):
- DockerContainerListNow: Lists every container Docker currently knows about right now (running or not), live – not a stored snapshot, over DockerProxyClient/the Jube.Monitoring sidecar’s own /proxy endpoint, never docker.sock directly. Read-only. (permissions from DockerAdministratorService.RequiredPermissions)
- DockerContainerMetricList: Lists per-minute, per-container Docker metrics (state, health, restart count, CPU/memory/network/block-io usage, pids) captured by the Jube.Monitoring sidecar, across every host, most recent first, capped at 100000, with optional date-range… (permissions from DockerContainerMetricService.RequiredPermissions)
- DockerEventList: Lists Docker Engine lifecycle events on the host (container die/kill/oom/restart/health_status, network/volume/image events, docker exec audit trail), most recent first, capped at 100000, with optional date-range and substring search (Event… (permissions from DockerEventService.RequiredPermissions)
- DockerHostInfoNow: Reads the Docker host’s own live info/version (container/image counts, CPU, memory, kernel, OS, Docker/API version) right now, over the same sidecar proxy path as DockerContainerListNow. Read-only. ( permissions from DockerAdministratorService.RequiredPermissions)
- DockerHostMetricList: Lists per-minute, per-host Docker host metrics (container/image counts, daemon version, image/layer/container/volume/build-cache disk usage) captured by the Jube.Monitoring sidecar, across every host, most recent first, capped at 100000, wi… (permissions from DockerHostMetricService.RequiredPermissions)
- HAProxyReachabilityProbeList: Lists per-minute synthetic HAProxy reachability probes. Landlord only. ( permissions from HaProxyReachabilityProbeService.RequiredPermissions)
- HAProxyServerStatusList: Lists per-minute HAProxy backend server status snapshots. Landlord only. (permissions from HaProxyServerStatusService.RequiredPermissions)
- OverlayNetworkTaskDriftList: Lists per-minute overlay network and DNS reconciliation checks. Landlord only. ( permissions from OverlayNetworkTaskDriftService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Etcd and Patroni Cluster State (`etcd-and-patroni-cluster-state`)

Hierarchy Placement: Level 3, child of Platform Reliability
Semantic Persona: Owns the etcd and patroni cluster state slice end to end, which is small enough that reads and writes stay with one agent.
Semantic Routing Keywords: etcd, cluster, event, member, status, patroni
Toolbox (Max 15):
- EtcdClearAlarm: Clears every currently-raised etcd alarm (NOSPACE/CORRUPT) on the given endpoint; acknowledges the flag, does not fix whatever raised it. Write — confirmation required (permissions from EtcdAdministratorService.RequiredPermissions)
- EtcdClusterEventList: Lists captured etcd cluster events (LeaderChanged, HealthChanged, AlarmRaised, AlarmCleared), detected by diffing consecutive one-minute EtcdMemberStatus samples, most recent first, capped at
  1. (permissions from EtcdClusterEventService.RequiredPermissions)
- EtcdClusterStatusNow: Reads etcd’s current member status, health, alarms and version right now over its own REST API, for every discoverable endpoint – not the background sampler’s last-stored row. Read-only. (permissions from EtcdAdministratorService.RequiredPermissions)
- EtcdDefragment: Defragments one etcd member’s own backend database file to reclaim freed space; briefly blocks that member, so call it on one member at a time, never the whole cluster at once. Write — confirmation required (permissions from EtcdAdministratorService.RequiredPermissions)
- EtcdHashKv: Reads one etcd member’s own keyspace consistency hash at its current revision, for comparing across members to catch a diverged member – two members at the same revision reporting different hashes have disagreed about the keyspace. Read-only. (permissions from EtcdAdministratorService.RequiredPermissions)
- EtcdMaintenanceStatus: Reads one etcd member’s raft term/index, current believed leader, and backend database size both total and in-use – the figure to check before running EtcdDefragment, rather than running it blind. Read-only. (permissions from EtcdAdministratorService.RequiredPermissions)
- EtcdMemberList: Reads etcd’s own authoritative membership list from the given endpoint – member id, name, peer/client URLs, and voting vs learner status. Any one reachable member answers for the whole cluster, since membership is Raft-replicated state. Read-only. (permissions from EtcdAdministratorService.RequiredPermissions)
- EtcdMemberStatusList: Lists per-minute etcd cluster member snapshots, one row per discovered/configured member, most recent first, capped at 100000. (permissions from EtcdMemberStatusService.RequiredPermissions)
- PatroniClusterEventList: Lists captured Patroni cluster events: RoleChanged/StateChanged/TimelineChanged (detected by diffing consecutive one-minute PatroniMemberStatus samples), and Failover (read directly from Patroni’s own DCS ‘history’ key in etcd, with Reason/… (permissions from PatroniClusterEventService.RequiredPermissions)
- PatroniClusterStatusNow: Reads every discoverable Patroni member’s status right now over its own REST API – role, state, timeline, lag, pending restart – not the background sampler’s last-stored row. Read-only. ( permissions from PatroniAdministratorService.RequiredPermissions)
- PatroniConfigGet: Reads a Patroni member’s own resolved dynamic configuration (DCS-stored config merged with whatever is local to that node) as plain JSON text. Read-only. (permissions from PatroniAdministratorService.RequiredPermissions)
- PatroniMemberStatusList: Lists per-minute Patroni-managed Postgres member snapshots, one row per discovered/configured member, most recent first, capped at 100000. (permissions from PatroniMemberStatusService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Postgres Observability (`postgres-observability`)

Hierarchy Placement: Level 3, child of Platform Reliability
Semantic Persona: Owns the postgres observability slice end to end, which is small enough that reads and writes stay with one agent.
Semantic Routing Keywords: postgres, activity, index, statistics, log, entry, metric, replication, status, statement, table
Toolbox (Max 15):
- PostgresActivityList: Lists every backend currently known to PostgreSQL (pg_stat_activity), live – not a stored snapshot, so results reflect only the instant this is called. (permissions from PostgresActivityService.RequiredPermissions)
- PostgresIndexStatisticsList: Lists live per-index Postgres statistics for every user index (scan counts, size, unique/primary-key/unused flags), queried live – not stored history, not scoped to any tenant. (permissions from PostgresIndexStatisticsService.RequiredPermissions)
- PostgresLockList: Lists every lock PostgreSQL currently knows about (pg_locks), live. Ungranted rows – a backend waiting on another one – sort first. Complements PostgresActivityList’s BlockedByPids (who is blocked) with what is actually being contended (which relation, which lock mode). Optional case-insensitive substring search against username, query text or relation name. Read-only. (permissions from PostgresLockService.RequiredPermissions)
- PostgresLogEntryList: Lists Postgres’s own server log lines (checkpoints, connection issues, deadlocks, crashes, slow autovacuum, …), most recent first, capped at 100000, with optional date-range and substring search (Message/Level) filters. (permissions from PostgresLogEntryService.RequiredPermissions)
- PostgresMetricList: Lists per-minute Postgres metric snapshots across every node, most recent first, capped at 100000, with optional date-range and substring search (Instance) filters, not scoped to any tenant. (permissions from PostgresMetricService.RequiredPermissions)
- PostgresReplicationStatusList: Lists per-minute pg_stat_replication snapshots, one row per connected standby, most recent first, capped at 100000, with optional date-range and substring search (UserName/ApplicationName/ClientAddress/State/SyncState) filters, not scoped t… (permissions from PostgresReplicationStatusService.RequiredPermissions)
- PostgresStatementStatisticsList: Lists per-query-shape aggregate statistics from pg_stat_statements – calls, rows, execution time min/mean/max/stddev, cache-hit vs disk-read bytes, temp-file spill bytes and WAL bytes, accumulated across all history since the extension was… (permissions from PostgresStatementStatisticsService.RequiredPermissions)
- PostgresTableStatisticsList: Lists live per-table Postgres statistics for every user table (scans, row estimates, sizes, vacuum/analyze history, a LikelyMissingIndex heuristic), queried live – not stored history, not scoped to any tenant. (permissions from PostgresTableStatisticsService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Redis Observability (`redis-observability`)

Hierarchy Placement: Level 3, child of Platform Reliability
Semantic Persona: Owns the redis observability slice end to end, which is small enough that reads and writes stay with one agent.
Semantic Routing Keywords: redis, call, counter, connection, event, multiplexer, metric, sentinel, status, slow, operation
Toolbox (Max 15):
- RedisCallCounterList: Lists per-minute Redis call counters – one row per distinct “call” label (a Jube.Cache repository method, e.g. (permissions from RedisCallCounterService.RequiredPermissions)
- RedisConnectionEventList: Lists captured CacheService ConnectionMultiplexer events (ConnectionFailed, ConnectionRestored, ErrorMessage, InternalError, ConfigurationChanged, ConfigurationChangedBroadcast), most recent first, capped at 100000. (permissions from RedisConnectionEventService.RequiredPermissions)
- RedisConnectionMultiplexerMetricList: Lists per-minute snapshots of the application’s own live Redis connection health and queue depth across every node, most recent first, capped at 100000, with optional date-range and substring search (Instance/ClientName) filters, not scoped… (permissions from RedisConnectionMultiplexerMetricService.RequiredPermissions)
- RedisKeyActivityList: Samples up to 100 keys from the live Redis keyspace (SCAN, never KEYS) and reports each one’s type, idle time, and – only under an LFU maxmemory-policy – its approximate access frequency. The 100-key cap is fixed. Optional glob pattern narrows the sample. For spotting a hot/cold key design problem, never for enumerating the keyspace. Read-only. (permissions from RedisKeyActivityService.RequiredPermissions)
- RedisMetricList: Lists per-minute Redis metric snapshots across every node, most recent first, capped at 100000, with optional date-range and substring search (Instance) filters, not scoped to any tenant. (permissions from RedisMetricService.RequiredPermissions)
- RedisSentinelCheckQuorum: Reads whether enough Sentinels currently agree a failover of the given master could actually succeed right now – the real pre-flight check before reaching for RedisSentinelFailover. Read-only. ( permissions from RedisSentinelAdministratorService.RequiredPermissions)
- RedisSentinelClusterStatusNow: Reads every master/slave/sentinel Sentinel currently reports right now over its own protocol, not the background sampler’s last-stored row, which can be up to a minute stale. Read-only. ( permissions from RedisSentinelAdministratorService.RequiredPermissions)
- RedisSentinelEventList: Lists captured Redis Sentinel pub/sub events (every channel Sentinel publishes – +sdown, -sdown, +odown, +switch-master, +failover-state, +slave, +sentinel, and more), most recent first, capped at 100000, with optional date-range and subst… (permissions from RedisSentinelEventService.RequiredPermissions)
- RedisSentinelFailover: Forces a manual failover of the given master right now – the action Sentinel’s own design exists to let an operator trigger under a real incident. Write — confirmation required (permissions from RedisSentinelAdministratorService.RequiredPermissions)
- RedisSentinelGetMasterAddress: Reads Sentinel’s own current, authoritative address for the given master, as “host:port”. Read-only. (permissions from RedisSentinelAdministratorService.RequiredPermissions)
- RedisSentinelReset: Forgets Sentinel’s current state of every master matching the given glob pattern, forcing Sentinel to rediscover it from scratch; self-healing and non-destructive to data but still a real mutating call. Write — confirmation required (permissions from RedisSentinelAdministratorService.RequiredPermissions)
- RedisSentinelStatusList: Lists per-minute Redis Sentinel topology snapshots (SENTINEL MASTERS/SLAVES/SENTINELS), one row per master/slave/sentinel entity, most recent first, capped at 100000. (permissions from RedisSentinelStatusService.RequiredPermissions)
- RedisSlowOperationList: Lists Redis SLOWLOG entries across every node, most recent first, capped at 100000, with optional date-range and substring search (Command/CommandName/KeyName/ClientAddress/ClientName) filters, not scoped to any tenant. (permissions from RedisSlowOperationService.RequiredPermissions)
Downstream Dependencies: none (leaf)

Fitting this to the runtime

The tree above is a design, not a wiring. Four things in Jube.AskJooby decide whether it can execute as drawn:

RunContext.MaxDepth defaults to 3 and AgentFactory only offers a delegation tool while depth < context.MaxDepth, so root to domain to supervisor to leaf is exactly the budget and there is no room for a fifth tier; the nine slices that reduced to a single leaf are attached straight to their domain rather than kept behind a supervisor that would route to one child, which also buys back a hop for them.
RunContext.MaxFanOut defaults to 8 and SpawnAgentTool rejects a single spawn call that exceeds it, which is a cap on how many specialists are asked at once and not on how many an agent may choose between; no roster here exceeds nine, counting cross-domain reach, and model-admin and model-support are separate domains specifically so that the slice rosters stay within eight before any cross-domain edge is added.
AgentToolDefinition.RequiresSnapshot is already derived from the owning operation being Write or Delete, and SnapshotOnWriteAiFunction wraps those tools; the 174 tools flagged above are that same set, so the flag is a restatement of what the runtime already computes and must not be reintroduced as a second, hand-maintained source of truth.
AgentDefinitions.IsPermitted checks RequiresLandlord before it considers RequiredPermissions, because 49 services are landlord-only with an empty permission list; every permission shown above is read from the owning service’s own RequiredPermissions and is never retyped here, so a service that changes its guard changes this tree with it.

Each agent’s toolbox is the authority for what it can call. Nothing above grants a tool that the catalogue does not already carry, and the eighteen operations the catalogue withholds on purpose stay withheld: they are absent from this tree by the same decision.

Agent hierarchy over the Jube service layer

Shape

Cross-domain reach

Root

Jooby

Access Control and Tenancy (access-admin)

Access Control and Tenancy

Entitlement Supervisor

Role Registry Inspector (role-registry-inspector)

Role Registry Test Bench (role-registry-test-bench)

Role Registry Author (role-registry-author)

Role Registry Permission Inspector (role-registry-permission-inspector)

Role Registry Permission Test Bench (role-registry-permission-test-bench)

Role Registry Permission Author (role-registry-permission-author)

Permission Specification (permission-specification)

Identity Supervisor

User Registry Inspector (user-registry-inspector)

User Registry Test Bench (user-registry-test-bench)

User Registry Author (user-registry-author)

User Observability (user-observability)

Tenancy Supervisor

Tenant Registry Inspector (tenant-registry-inspector)

Tenant Registry Test Bench (tenant-registry-test-bench)

Tenant Registry Author (tenant-registry-author)

Assistant Conversation Administration (assistant-admin)

Assistant Conversation Administration

Ask Jooby Observability (ask-jooby-observability)

Case Investigation Desk (case-analyst)

Case Investigation Desk

Investigation Supervisor

Case Record and Files (case-record-and-files)

Case Lookup and Notes (case-lookup-and-notes)

Case Journal and Events (case-journal-and-events)

Activation Watcher and Archive (activation-watcher-and-archive)

Sanctions Entry Source (sanctions-entry-source)

Session Case Observability (session-case-observability)

Case Workflow Governance (case-workflow-admin)

Case Workflow Governance

Definition Supervisor

Case Workflow Inspector (case-workflow-inspector)

Case Workflow Test Bench (case-workflow-test-bench)

Case Workflow Author (case-workflow-author)

Entitlement Supervisor

Action and Display Role Grants (action-and-display-role-grants)

Filter and Form Role Grants (filter-and-form-role-grants)

Macro and Workflow Role Grants (macro-and-workflow-role-grants)

Status and XPath Role Grants (status-and-xpath-role-grants)

Interaction Actions Supervisor

Case Workflow Action Inspector (case-workflow-action-inspector)

Case Workflow Action Test Bench (case-workflow-action-test-bench)

Case Workflow Action Author (case-workflow-action-author)

Case Workflow Display Inspector (case-workflow-display-inspector)

Case Workflow Display Test Bench (case-workflow-display-test-bench)

Case Workflow Display Author (case-workflow-display-author)

Interaction Forms Supervisor

Case Workflow Form Inspector (case-workflow-form-inspector)

Case Workflow Form Test Bench (case-workflow-form-test-bench)

Case Workflow Form Author (case-workflow-form-author)

Form Entries and Macro Executions (form-entries-and-macro-executions)

Interaction Macros Supervisor

Case Workflow Macro Inspector (case-workflow-macro-inspector)

Case Workflow Macro Test Bench (case-workflow-macro-test-bench)

Case Workflow Macro Author (case-workflow-macro-author)

Routing Extraction Supervisor

Case Workflow Filter Inspector 1 (case-workflow-filter-inspector-1)

Case Workflow Filter Inspector 2 (case-workflow-filter-inspector-2)

Case Workflow Filter Test Bench (case-workflow-filter-test-bench)

Case Workflow Filter Author (case-workflow-filter-author)

Case Workflow X Path Inspector (case-workflow-x-path-inspector)

Case Workflow X Path Test Bench (case-workflow-x-path-test-bench)

Case Workflow X Path Author (case-workflow-x-path-author)

Routing Status Supervisor

Case Workflow Status Inspector 1 (case-workflow-status-inspector-1)

Case Workflow Status Inspector 2 (case-workflow-status-inspector-2)

Case Workflow Status Test Bench (case-workflow-status-test-bench)

Case Workflow Status Author (case-workflow-status-author)

Case Workflow Priority (case-workflow-priority)

Analytics Reporting (data-analyst)

Analytics Reporting

Visualisation Definition Supervisor

Access Control and Tenancy (`access-admin`)

Role Registry Inspector (`role-registry-inspector`)

Role Registry Test Bench (`role-registry-test-bench`)

Role Registry Author (`role-registry-author`)

Role Registry Permission Inspector (`role-registry-permission-inspector`)

Role Registry Permission Test Bench (`role-registry-permission-test-bench`)

Role Registry Permission Author (`role-registry-permission-author`)

Permission Specification (`permission-specification`)

User Registry Inspector (`user-registry-inspector`)

User Registry Test Bench (`user-registry-test-bench`)

User Registry Author (`user-registry-author`)

User Observability (`user-observability`)

Tenant Registry Inspector (`tenant-registry-inspector`)

Tenant Registry Test Bench (`tenant-registry-test-bench`)

Tenant Registry Author (`tenant-registry-author`)

Assistant Conversation Administration (`assistant-admin`)

Ask Jooby Observability (`ask-jooby-observability`)

Case Investigation Desk (`case-analyst`)

Case Record and Files (`case-record-and-files`)

Case Lookup and Notes (`case-lookup-and-notes`)

Case Journal and Events (`case-journal-and-events`)

Activation Watcher and Archive (`activation-watcher-and-archive`)

Sanctions Entry Source (`sanctions-entry-source`)

Session Case Observability (`session-case-observability`)

Case Workflow Governance (`case-workflow-admin`)

Case Workflow Inspector (`case-workflow-inspector`)

Case Workflow Test Bench (`case-workflow-test-bench`)

Case Workflow Author (`case-workflow-author`)

Action and Display Role Grants (`action-and-display-role-grants`)

Filter and Form Role Grants (`filter-and-form-role-grants`)

Macro and Workflow Role Grants (`macro-and-workflow-role-grants`)

Status and XPath Role Grants (`status-and-xpath-role-grants`)

Case Workflow Action Inspector (`case-workflow-action-inspector`)

Case Workflow Action Test Bench (`case-workflow-action-test-bench`)

Case Workflow Action Author (`case-workflow-action-author`)

Case Workflow Display Inspector (`case-workflow-display-inspector`)

Case Workflow Display Test Bench (`case-workflow-display-test-bench`)

Case Workflow Display Author (`case-workflow-display-author`)

Case Workflow Form Inspector (`case-workflow-form-inspector`)

Case Workflow Form Test Bench (`case-workflow-form-test-bench`)

Case Workflow Form Author (`case-workflow-form-author`)

Form Entries and Macro Executions (`form-entries-and-macro-executions`)

Case Workflow Macro Inspector (`case-workflow-macro-inspector`)

Case Workflow Macro Test Bench (`case-workflow-macro-test-bench`)

Case Workflow Macro Author (`case-workflow-macro-author`)

Case Workflow Filter Inspector 1 (`case-workflow-filter-inspector-1`)

Case Workflow Filter Inspector 2 (`case-workflow-filter-inspector-2`)

Case Workflow Filter Test Bench (`case-workflow-filter-test-bench`)

Case Workflow Filter Author (`case-workflow-filter-author`)

Case Workflow X Path Inspector (`case-workflow-x-path-inspector`)

Case Workflow X Path Test Bench (`case-workflow-x-path-test-bench`)

Case Workflow X Path Author (`case-workflow-x-path-author`)

Case Workflow Status Inspector 1 (`case-workflow-status-inspector-1`)

Case Workflow Status Inspector 2 (`case-workflow-status-inspector-2`)

Case Workflow Status Test Bench (`case-workflow-status-test-bench`)

Case Workflow Status Author (`case-workflow-status-author`)

Case Workflow Priority (`case-workflow-priority`)

Analytics Reporting (`data-analyst`)

Visualisation Registry Inspector (`visualisation-registry-inspector`)

Visualisation Registry Test Bench (`visualisation-registry-test-bench`)

Visualisation Registry Author (`visualisation-registry-author`)

Visualisation Role Grants and Execution (`visualisation-role-grants-and-execution`)

Visualisation Registry Parameter Inspector (`visualisation-registry-parameter-inspector`)

Visualisation Registry Parameter Test Bench (`visualisation-registry-parameter-test-bench`)

Visualisation Registry Parameter Author (`visualisation-registry-parameter-author`)

Visualisation Registry Datasource Inspector (`visualisation-registry-datasource-inspector`)

Visualisation Registry Datasource Test Bench (`visualisation-registry-datasource-test-bench`)

Visualisation Registry Datasource Author (`visualisation-registry-datasource-author`)

Datasource and Parameter Role Grants (`datasource-and-parameter-role-grants`)

Detection Model Authoring (`model-admin`)

Entity Analysis Model Abstraction Calculation Inspector (`entity-analysis-model-abstraction-calculation-inspector`)

Entity Analysis Model Abstraction Calculation Test Bench (`entity-analysis-model-abstraction-calculation-test-bench`)

Entity Analysis Model Abstraction Calculation Author (`entity-analysis-model-abstraction-calculation-author`)

Entity Analysis Model Abstraction Rule Inspector (`entity-analysis-model-abstraction-rule-inspector`)

Entity Analysis Model Abstraction Rule Test Bench (`entity-analysis-model-abstraction-rule-test-bench`)

Entity Analysis Model Abstraction Rule Author (`entity-analysis-model-abstraction-rule-author`)

Entity Analysis Model Abstraction Rule Approvals (`entity-analysis-model-abstraction-rule-approvals`)

Entity Analysis Model Ttl Counter Inspector (`entity-analysis-model-ttl-counter-inspector`)

Entity Analysis Model Ttl Counter Test Bench (`entity-analysis-model-ttl-counter-test-bench`)

Entity Analysis Model Ttl Counter Author (`entity-analysis-model-ttl-counter-author`)