Run failure analysis on a trace or minimal input dataset
Valiqor API key. Pass as Bearer token: Authorization: Bearer vq-xxxxxxxxxxxxxxxxxxxx
Request for failure analysis.
Trace ID for full trace analysis
Dataset items for minimal input analysis
Project name for organization
App type hint: 'rag', 'agent', 'agentic_rag', 'generic_llm'. Auto-detected if not provided.
Run evaluation metrics alongside failure analysis
Run security audit alongside failure analysis
Fetch and attach AST scan data if available (auto-enabled for trace inputs)
User-specified eval metrics that must run
User-specified security categories that must run
Filter to specific failure subcategories. If None, runs all applicable.
Filter to specific failure buckets (e.g., 'hallucination_grounding', 'instruction_compliance'). All subcategories within the specified buckets will be included. If both buckets and subcategories are provided, they are merged (union).
OpenAI API key for LLM judges
Successful Response
Full failure analysis response.
'completed', 'processing', 'failed'
'minimal' or 'full'
'trace', 'json', or 'csv'
Detected or specified app type
Summary of failure analysis results.
Original input items with per-item failure stats
Evaluation metric scores
ID of linked evaluation run
Security category flags
ID of linked security audit batch
Subcategories that were analyzed
Subcategories skipped (not applicable)