Outcome Evidence

The fifth coordination primitive. Proof fields, not narration.

What is Evidence?

Evidence is the fifth of CueAPI's five coordination primitives: structured proof fields attached to an outcome. Not narration. When your agent posts a tweet, sends an email, or creates a Stripe charge, you attach the external ID and URL as evidence. Evidence is how CueAPI verifies a handoff really happened, not just that someone claimed it did.

Evidence Fields

Include these fields when reporting an outcome via POST /v1/executions/{id}/outcome:

FieldTypeDescription
external_idstringExternal system ID (e.g., tweet:123456, charge:ch_abc)
result_urlstringURL to the created resource
result_typestringType of result: tweet, email, stripe_charge, file
summarystringHuman-readable description (max 500 chars)
artifactsarrayList of created resource IDs or URLs

Example

bash
curl -X POST https://api.cueapi.ai/v1/executions/{id}/outcome \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "success": true,
    "result": "Morning briefing posted",
    "external_id": "tweet:1234567890",
    "result_url": "https://twitter.com/user/status/1234567890",
    "result_type": "tweet",
    "summary": "Morning briefing: 3 key updates"
  }'

Querying by Evidence

Filter executions by evidence fields:

bash
# All executions with evidence
GET /v1/executions?has_evidence=true
 
# All tweet outcomes
GET /v1/executions?result_type=tweet
 
# Verification failures
GET /v1/executions?outcome_state=verification_failed
How do I know if my agent ran successfully?
Ctrl+K