Jump to the main Foxhound workflows quickly.
Healthy execution in 8.52s. Cost: $0.0406 across 3 LLM calls.
{
"workflow": "refund_and_return_resolution",
"environment": "production",
"issue_type": "refund",
"customer_tier": "self-serve",
"prompt_name": "refund-policy-check",
"prompt_version": 3,
"release_version": "2026.04.13",
"model_provider": "openai",
"model_name": "gpt-4o",
"ticket_id": "ticket_refund_clarification_request_avoided_a_false_denial_7_20_session_refund_clarification_07",
"language": "en",
"user_id": "customer_refund",
"agent_profile_id": "returns-resolution-copilot",
"agent_team": "Post-purchase Ops",
"agent_owner": "Support automation",
"agent_mission": "Handles refund, exchange, and return-eligibility decisions for self-serve customers.",
"story_label": "Refund clarification request avoided a false denial (7.20)",
"story_summary": "Returns Resolution Copilot avoided an incorrect refund denial on day 7 by asking for the missing evidence before deciding.",
"scenario_id": "generated_refund_clarification_needed",
"customer_intent": "Customer supplied partial return evidence and asked for a refund exception.",
"expected_resolution": "Request missing evidence before making a policy decision.",
"policy_result": "clarification_requested",
"quality_score": 0.95,
"severity_tag": "healthy",
"status_label": "Clarification path",
"replay_label": "Returns Resolution Copilot · Refund clarification request avoided a false denial (7.20)",
"operator_notes": "Case: Refund clarification request avoided a false denial (7.20)\nIntent: Customer supplied partial return evidence and asked for a refund exception.\nExpected resolution: Request missing evidence before making a policy decision.\nObserved policy result: clarification_requested\nStory summary: Returns Resolution Copilot avoided an incorrect refund denial on day 7 by asking for the missing evidence before deciding.\nAgent role: Returns Resolution Copilot\nOwner team: Post-purchase Ops",
"generated_fixture": true,
"generated_template": "refund_clarification_needed",
"cohort": "peak_load",
"operator_segment": "Post-purchase Ops",
"narrative_batch": "week_7",
"call_reason_detail": "refund_3"
}