Skip to content

Policy knowledge

M
All runs
Passedscheduled

Search returns the policy active on a requested date

1/1 sample passed1.7 s

Agent: Policy search API · JSON

Captured evidence

Sample evidence

1 sample
PassedSample 11.7 s
Scenario: A manager requests the US return policy that was active on a fixed date.

Test input

{
  "query": "standard return window",
  "region": "US",
  "asOf": "2025-08-01"
}

Agent response

{
  "policyId": "returns-us-v10",
  "title": "Standard returns — United States",
  "region": "US",
  "version": 10,
  "status": "effective",
  "effectiveAt": "2025-07-01",
  "returnWindowDays": 30
}

Pass condition results

3

  • Must include “returns-us-v10”

    Passed

    Required content was present.

  • Must include “2025-07-01”

    Passed

    Required content was present.

  • The JSON result clearly answers the standard-return query with a coherent effective-policy summary.

    Passed

    The result returns the expected US policy, version 10, effective status, and the correct effective date for the requested as-of date.

    Evaluation details
    {
  "minimumScore": 0.95,
  "model": "openai/gpt-5.6-luna",
  "promptRevision": "rubric-v1",
  "score": 1,
  "verdict": "scored"
}
Technical details
Started
Completed
Agent revision
policy-search-v5

Request envelope

{
  "actor": "store-manager",
  "agent": "policy-search",
  "checkId": "check_policy_current",
  "input": {
    "query": "standard return window",
    "region": "US",
    "asOf": "2025-08-01"
  },
  "projectId": "project_policy",
  "requestId": "b70b7eb2-31b9-48ef-bec1-c683cc3ea32e",
  "runId": "run_policy_current_latest",
  "sample": 1,
  "scenarioId": "us-return-policy-as-of-date",
  "schemaVersion": "1"
}

Response envelope

{
  "output": {
    "policyId": "returns-us-v10",
    "title": "Standard returns — United States",
    "region": "US",
    "version": 10,
    "status": "effective",
    "effectiveAt": "2025-07-01",
    "returnWindowDays": 30
  },
  "revision": "policy-search-v5",
  "schemaVersion": "1"
}
Technical details
Run ID
run_policy_current_latest
Check ID
check_policy_current
State
completed
Started
Completed