trajectories

Critique

LLM-side critique. Synchronous stub for now — enqueue a job and return a placeholder; phase-4 replaces this with a real deepseek call.

The shape returned matches what the production critique will return: a list of findings, each with a record_uuid anchor, severity, and body. UI can render these as auto-annotations the mentor can accept / convert into real annotations.

post/api/trajectories/{session_id}/critique

Path parameters

session_idstring required

Headers

authorizationstring nullable

Response

Successful Response

object required

Changes

No recorded changes to this endpoint across all 2 revisions of this API.