Skip to main content
GET
/
v1
/
analytics
/
failure-analysis
curl "https://api.playgent.com/v1/analytics/failure-analysis?agent_id=agent_abc123&period=30d" \
  -H "Authorization: Bearer your-api-key"
{
  "period": "30d",
  "total_failures": 156,
  "failure_categories": [
    {
      "category": "Hallucination",
      "count": 45,
      "percentage": 28.8,
      "example_run_ids": ["run_1", "run_2"],
      "common_patterns": [
        "Made up order numbers not in context",
        "Stated incorrect refund policy details"
      ],
      "suggested_fix": "Add explicit instruction: 'Only reference order numbers provided in the context'"
    },
    {
      "category": "Off-topic Response",
      "count": 32,
      "percentage": 20.5,
      "example_run_ids": ["run_3", "run_4"],
      "common_patterns": [
        "Discussed unrelated products",
        "Provided information not asked for"
      ],
      "suggested_fix": "Add guardrail: 'Stay focused on the customer's specific question'"
    },
    {
      "category": "Incomplete Response",
      "count": 28,
      "percentage": 17.9,
      "example_run_ids": ["run_5", "run_6"],
      "common_patterns": [
        "Did not address all parts of multi-part questions",
        "Missing follow-up information"
      ],
      "suggested_fix": "Add instruction: 'Ensure all parts of the user's question are addressed'"
    }
  ]
}

Documentation Index

Fetch the complete documentation index at: https://playgent.mintlify.app/llms.txt

Use this file to discover all available pages before exploring further.

Get intelligent analysis of test failures including categorization, common patterns, and actionable suggestions for improvement.
agent_id
string
Filter by agent ID (optional)
period
string
Time period to analyze (default: “30d”)
period
string
required
Analysis period
total_failures
integer
required
Total number of failures
failure_categories
array
required
Categorized failures with analysis
curl "https://api.playgent.com/v1/analytics/failure-analysis?agent_id=agent_abc123&period=30d" \
  -H "Authorization: Bearer your-api-key"
{
  "period": "30d",
  "total_failures": 156,
  "failure_categories": [
    {
      "category": "Hallucination",
      "count": 45,
      "percentage": 28.8,
      "example_run_ids": ["run_1", "run_2"],
      "common_patterns": [
        "Made up order numbers not in context",
        "Stated incorrect refund policy details"
      ],
      "suggested_fix": "Add explicit instruction: 'Only reference order numbers provided in the context'"
    },
    {
      "category": "Off-topic Response",
      "count": 32,
      "percentage": 20.5,
      "example_run_ids": ["run_3", "run_4"],
      "common_patterns": [
        "Discussed unrelated products",
        "Provided information not asked for"
      ],
      "suggested_fix": "Add guardrail: 'Stay focused on the customer's specific question'"
    },
    {
      "category": "Incomplete Response",
      "count": 28,
      "percentage": 17.9,
      "example_run_ids": ["run_5", "run_6"],
      "common_patterns": [
        "Did not address all parts of multi-part questions",
        "Missing follow-up information"
      ],
      "suggested_fix": "Add instruction: 'Ensure all parts of the user's question are addressed'"
    }
  ]
}