ouroboros_start_evaluate
Start a background evaluation of an execution artifact, returning a job ID immediately to prevent client timeouts. Later, poll job status and receive the verdict.
Instructions
Start an evaluation in the background and return a job ID immediately. Use this instead of ouroboros_evaluate when the three-stage pipeline (mechanical + semantic + optional consensus) is expected to exceed the MCP client tool-call timeout. Poll with ouroboros_job_status / ouroboros_job_wait and read the verdict via ouroboros_job_result. In plugin mode with auto_evolve disabled, evaluation is delegated to an OpenCode Task pane and job_id is None. With auto_evolve enabled, the parent keeps evaluation pollable so a rejection can start Ralph.
Input Schema
TableJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
| artifact | Yes | The execution output/artifact to evaluate | |
| session_id | Yes | The execution session ID to evaluate | |
| auto_evolve | No | Override execution.auto_evolve for this background call. When true, an explicitly rejected formal evaluation starts a bounded Ralph continuation loop. | |
| working_dir | No | Project root used to resolve Stage 1 mechanical verification commands and Stage 2 source-file visibility. Commands are read from .ouroboros/mechanical.toml; when the file is missing, the evaluator makes one AI detect call that inspects manifests (package.json, pyproject.toml, Cargo.toml, Makefile, ...) and authors the toml. Stage 1 skips every check when no toml is produced — it never guesses. | |
| seed_content | No | Original seed YAML for goal/constraints extraction | |
| artifact_type | No | Type of artifact: code, docs, config. Default: code | code |
| seed_handoff_id | No | Opaque parent-owned Seed handle for plugin evaluation | |
| trigger_consensus | No | Force Stage 3 consensus evaluation. Default: False | |
| acceptance_criteria | No | Multiple acceptance criteria for checklist evaluation. When two or more items are provided, each AC is evaluated independently and the results are aggregated into a pass/fail checklist (#366). Overrides acceptance_criterion. | |
| acceptance_criterion | No | Specific acceptance criterion to evaluate against |