Save agent answers or allowed preferences
Requires an authenticated user JWT. Generated from the production route registered in `internal/features/agents/inbound/setup.go`. At least one answer/preference change is required. Null clears a value. Only declared answer keys and typed preferences are accepted, with configured agent text screening. Stale revisions or a reused request_id with different input conflict. An exact replay returns current state, never rewinds a newer edit. Answer strings are bounded to 4000 UTF-8 bytes and total saved answers/preferences to 32 KiB.
/v1/agents/{id}/setupRequires an authenticated user JWT. Generated from the production route registered in internal/features/agents/inbound/setup.go. At least one answer/preference change is required. Null clears a value. Only declared answer keys and typed preferences are accepted, with configured agent text screening. Stale revisions or a reused request_id with different input conflict. An exact replay returns current state, never rewinds a newer edit. Answer strings are bounded to 4000 UTF-8 bytes and total saved answers/preferences to 32 KiB.
Authorization
bearerAuth Engine-issued API key or tenant-signed JWT.
In: header
Path Parameters
uuidRequest Body
application/json
TypeScript Definitions
Use the request body type in TypeScript.
Response Body
application/json
application/json
application/json
application/json
application/json
application/json
application/json
application/json
curl -X PATCH "https://example.com/v1/agents/497f6eca-6276-4993-bfeb-53cbbbba6f08/setup" \ -H "Content-Type: application/json" \ -d '{ "request_id": "266ea41d-adf5-480b-af50-15b940c2b846", "expected_revision": 0 }'{ "agent_id": "2b1e3b65-2c04-4fa2-a2d7-467901e98978", "revision": 0, "answers": { "property1": "string", "property2": "string" }, "preferences": { "property1": "string", "property2": "string" }, "conversation_id": "cc71b11a-25cd-4c2d-9950-df2cc38e3407", "role": { "role_id": "string", "edition": 1, "name": "string", "description": "string", "expertise": "string", "setup_instructions": "string", "requirements": [ { "key": "string", "kind": "answer", "label": "string", "prompt": "string", "shared": true, "tool": "string" } ], "preferences": [ { "key": "string", "label": "string", "kind": "text", "options": [ "string" ], "max_length": 1 } ], "setup_required": true, "blinq_id": "163ce100-8211-4522-93e1-e382b5984dbb", "blinq_version": 1, "presentation": { "color": "string", "avatar": "string", "availability": "available" }, "executor_mode": "internal", "core_workflow": [], "jobs": [] }, "shared_ready": true, "requirements": [ { "key": "string", "kind": "answer", "label": "string", "prompt": "string", "shared": true, "tool": "string", "status": "ready", "reason": "string" } ], "jobs": [ { "id": "string", "name": "string", "ready": true, "missing": [ "string" ] } ]}Read agent setup and live job readiness
Requires an authenticated user JWT. Generated from the production route registered in `internal/features/agents/inbound/setup.go`. Owner-only. Each job checks shared and job-specific answers, current tool grants/policy/availability and knowledge access. Failed checks remain unavailable. The role document is returned for trusted application adapters; clients can expose only the visible questions, choices and job summaries.
Begin or continue a saved agent setup chat
Requires an authenticated user JWT. Generated from the production route registered in `internal/features/agents/inbound/setup_chat.go`. A request_id-only body begins/resumes the real conversation without a model call. Natural replies require expected_revision and use governed tool-free completion to distinguish answers from clarification. Configured agent text policies apply; only screened owner text is saved. The reply and answers commit together. Completed retries return current history; uncertain claimed replies conflict without repeating model spend. pending_job identifies the router's saved request and current readiness.