Stream Skill Eval
Run the benchmark, streaming per-case progress (SSE): bootstrapping ->
start -> case_started / case_done (per case) -> summary. Feeds
the 'Proven on' surface case by case. strict=true runs the Phase 2 gate.
Path Parameters
skill_id*Skill Id
Query Parameters
strict?Strict
Default
falseResponse Body
application/json
application/json
curl -X GET "https://example.com/skills/string/eval/stream"null{ "detail": [ { "loc": [ "string" ], "msg": "string", "type": "string", "input": null, "ctx": {} } ]}