List Evaluation Runs
Database selection
Use an evaluation-capable database registered with AgentOS or its components. The example API server registers SQLite through its agent. When several database IDs are available, supply db_id; selecting table also requires db_id. These values select registered storage, not an arbitrary database connection. Authentication and configured user scoping apply to the stored evaluation records.
The response contains both data and meta. The generated example below abbreviates this envelope. For an empty first page, the shape is:
{"data":[],"meta":{"page":1,"limit":20,"total_count":0,"total_pages":0}}eval_types accepts comma-separated evaluation types. Unsupported values return 422.
/eval-runsRetrieve paginated evaluation runs with filtering and sorting options. Filter by agent, team, workflow, model, or evaluation type.
Authorization
HTTPBearer In: header
Query Parameters
Agent ID
Team ID
Workflow ID
Model ID
Filter type
Number of eval runs to return
20Page number
1Field to sort by
"created_at"Sort order (asc or desc)
"desc"The ID of the database to use
The database table to use
Comma-separated eval types (accuracy,agent_as_judge,performance,reliability)
Response Body
application/json
application/json
application/json
application/json
application/json
application/json
curl --request GET 'https://example.com/eval-runs'{ "data": [ { "id": "a03fa2f4-900d-482d-afe0-470d4cd8d1f4", "agent_id": "basic-agent", "model_id": "gpt-4o", "model_provider": "OpenAI", "name": "Test ", "eval_type": "reliability", "eval_data": { "eval_status": "PASSED", "failed_tool_calls": [], "passed_tool_calls": [ "multiply" ] }, "eval_input": { "expected_tool_calls": [ "multiply" ] }, "created_at": "2025-08-27T15:41:59Z", "updated_at": "2025-08-27T15:41:59Z" } ]}{ "detail": "string", "error_id": "string", "error_type": "string"}{ "detail": "string", "error_id": "string", "error_type": "string"}{ "detail": "string", "error_id": "string", "error_type": "string"}{ "detail": "string"}{ "detail": "string", "error_id": "string", "error_type": "string"}