Evaluations

Trigger a skill eval run

Runs a behavioral eval for a published skill by scoring its test cases against the latest completed sandbox outputs, then persists a graded result with pass rate and improvement. Requires authentication; only the skill's author may run evals. The skill must be published and have at least one test case.

post/api/v1/skills/{slug}/evals

Request

  • Base URL: https://api.skills-hub.ai
  • URL: https://api.skills-hub.ai/api/v1/skills/{slug}/evals
  • Auth: one of:
    • HTTP bearer
    • API key in header Authorization

Path parameters

slugstring required

Response

Successful response. The returned fields are described in this endpoint's summary/description; typed per-field response schemas are being added incrementally.

idstring required
skillIdstring required
versionIdstring nullable required
statusstring required
passRatenumber nullable required
improvementnumber nullable required
totalCasesinteger required
passedCasesinteger required
durationMsnumber nullable required
createdAtstring date-time required

Changes

No recorded changes to this operation across all 1 revision of this API.