Skip to main content
v1

Audit Trace API

Endpoint​

GET /api/v1/conversation/audit/{conversationId}/trace

Response shape​

Trace response
JSON
{
  "conversationId": "9bf7540a-b129-4685-a120-730e8a0cb94b",
  "steps": [
    {
      "step": "IntentResolutionStep",
      "stepClass": "com.github.salilvnair.convengine.engine.steps.IntentResolutionStep",
      "status": "OK",
      "startedAt": "2026-02-14T10:22:01.102Z",
      "endedAt": "2026-02-14T10:22:01.147Z",
      "durationMs": 45,
      "error": null,
      "stages": []
    }
  ],
  "stages": [
    {
      "auditId": 12031,
      "stage": "STEP_ENTER",
      "createdAt": "2026-02-14T10:22:01.102Z",
      "payload": {
        "step": "IntentResolutionStep"
      }
    }
  ]
}

Why this endpoint exists​

  • separates timeline from raw audit payload parsing
  • easier UI rendering for runbooks and debug tools
  • complements ce_conversation_history, which stores normalized user/AI turns even when audit persistence runs deferred
  • preserves both domain and technical stages
Best practice

Use both endpoints together: /audit/{id} for raw forensic payload, /audit/{id}/trace for step-flow visualization.