Agent Run Context

Operations of the LLM Registry API. Paths are relative to your deployment's API host.

POST /api/llm-registry-api/agent-run-context/render

Render the unified xema workspace context for an agent invocation.

Composes Xema's AGENTS.md and system overlay (placed in the runtime-config mount), /workspace/context.json, and a mount plan from the resolved agent snapshot (llm-registry) plus the resolved deliverable spec (deliverable-specs-api). Called by the xema/agent@v1 activity (and indirectly by xema/review@v3, whose draft step dispatches xema/agent@v1 per iteration) BEFORE driving the OpenCode session. Deterministic: same inputs → identical bytes → stable hashes; Temporal retries reuse the same render.

Authentication: bearer token (Authorization: Bearer <token>)

Request body: application/json RenderAgentRunContextRequestDto

StatusDescriptionBody
200application/json RenderAgentRunContextResponseDtoDataEnvelope
400application/problem+json ProblemDetailsDto
401application/problem+json ProblemDetailsDto
403application/problem+json ProblemDetailsDto
404application/problem+json ProblemDetailsDto
409application/problem+json ProblemDetailsDto
422application/problem+json ProblemDetailsDto
500application/problem+json ProblemDetailsDto