Start a chat test run

POST

https://api.levrage.ai/v1/agents/:agent_id/test-runs

Run up to 8 saved scenarios against the agent: a simulated caller says the scenario’s lines and every must / must-not rule is graded, with the agent’s own words as proof. Returns at once with a run id; read progress and results with GET /v1/test-runs/{run_id}. Uses the monthly test allowance.

Authentication

AuthorizationBearer

Bearer authentication of the form Bearer <token>, where token is your auth token.

Path parameters

agent_idstringRequired

Request

This endpoint expects an object.

scenario_idslist of stringsRequired

instructions_overridestring or nullOptional<=60000 characters

Run against these instructions instead of the saved ones, to prove a fix first. No report is saved.

save_reportboolean or nullOptional

Ignored with instructions_override (never saved)

Response

Successful Response

Errors

422

Unprocessable Entity Error