---
title: "Run experimental agent evals"
method: POST
path: "/api/agent/evals/run"
tags: ["Agent"]
---

# Run experimental agent evals

`POST /api/agent/evals/run`

## Request body

- AgentEvalRunRequest
  - `runId` string — Optional run ID to score. If omitted, recent runs for the current user/org are scored.
  - `limit` integer — Maximum recent runs to score when runId is omitted

## Response `200`

Eval run completed

- object
  - `status` 'completed', required
  - `results` AgentEvalResult[], required
    - `runId` string, required
    - `sessionId` string — Chat session that owns this agent run.
    - `status` 'running' | 'completed' | 'failed' | 'cancelled', required
    - `createdAt` string, date-time, required
    - `updatedAt` string, date-time, required
    - `evidenceCount` integer, required
    - `approvalCount` integer, required
    - `unresolvedApprovalCount` integer, required
    - `finalReportPresent` boolean, required
    - `scores` AgentEvalScores, required
      - `evidenceCoverage` integer, required
      - `approvalCompliance` integer, required
      - `finalReportCompleteness` integer, required
      - `rcaQuality` integer, required
      - `hallucinationRisk` integer, required
      - `overall` integer, required
    - `warnings` string[]

## Other responses

- `403` — Forbidden (insufficient permissions or RBAC violation)
- `404` — Resource not found

---

[API](https://skmtc.net/consensys/apis/ask-o11y-ai-powered-observability-assistant-api.md) · [All operations](https://skmtc.net/consensys/apis/ask-o11y-ai-powered-observability-assistant-api/llms.txt) · [OpenAPI document](https://skmtc-service-staging.skmtc.workers.dev/v1/apis/consensys/ask-o11y-ai-powered-observability-assistant-api/versions/3ccc03eccaef/schema)