---

name: pipeline-insights description: AI-powered pipeline diagnostics. Explains why faces weren't detected, tracked, embedded, or clustered using OpenAI GPT-4o analysis. Use when debugging pipeline issues, reviewing unclustered tracks, or understanding why quality gates rejected faces.

Pipeline Insights Skill

Use this skill to get AI-powered explanations for pipeline failures.

When to Use

• Track wasn't clustered and you don't know why
• Faces show "skipped" but the reason isn't clear
• Quality gate rejected faces you expected to pass
• Need to understand which config threshold to adjust
• Batch diagnosing unclustered tracks

Quick Start

Diagnose a Single Track

Copycurl -X POST "http://localhost:8000/diagnostics/episodes/{ep_id}/diagnose_track/{track_id}"

Diagnose All Unclustered Tracks

Copycurl -X POST "http://localhost:8000/diagnostics/episodes/{ep_id}/diagnose_unclustered" \
  -H "Content-Type: application/json" \
  -d '{"max_tracks": 50}'

Check Service Status

Copycurl "http://localhost:8000/diagnostics/status"

Understanding Results

Stage Progression

Each track progresses through these stages:

Common Failure Patterns

"embedding:min_blur_score"

Cause: Face too blurry (motion blur, out of focus)

Check: max_blur_score in diagnostic vs threshold (default: 18.0)

Fix: Lower config/pipeline/faces_embed_sampling.yaml:

Copyquality_gating:
  min_blur_score: 10.0  # default is 18.0

"clustering:min_identity_sim"

Cause: Embedding didn't match any existing cluster

Check: nearest_cluster_sim in diagnostic

Fix: Lower config/pipeline/clustering.yaml:

Copycluster_thresh: 0.44  # default is 0.52
min_identity_sim: 0.40  # default is 0.45

"detection:confidence"

Cause: Face detection confidence too low

Fix: Lower config/pipeline/detection.yaml:

Copyconfidence_th: 0.40  # default is 0.50

Config Reference

Embedding Thresholds

File: config/pipeline/faces_embed_sampling.yaml

Clustering Thresholds

File: config/pipeline/clustering.yaml

Diagnostic Output Example

Copy{
  "track_id": 59,
  "stage_reached": "tracked",
  "stage_failed": "embedded",
  "raw_data": {
    "faces_count": 1,
    "faces_skipped": 1,
    "skip_reasons": ["blurry:10.8"],
    "max_blur_score": 10.78,
    "embedding_generated": false
  },
  "ai_analysis": {
    "explanation": "Track 59 has only 1 face, and that face was too blurry (score 10.78) to generate an embedding. The minimum blur threshold is 18.0.",
    "root_cause": "Single-frame track with blur below quality gate",
    "blocked_by": "embedding:min_blur_score",
    "suggested_fixes": [
      "Lower min_blur_score to 10.0",
      "Manually assign track in Faces Review UI"
    ],
    "config_changes": [{
      "file": "config/pipeline/faces_embed_sampling.yaml",
      "key": "quality_gating.min_blur_score",
      "current": 18.0,
      "suggested": 10.0,
      "reason": "Allow blurrier faces for short tracks"
    }]
  }
}

Saved Reports

Diagnostic reports are saved to:

Copydata/manifests/{ep_id}/diagnostics/
├── track_0059_diagnostic.json
├── track_0127_diagnostic.json
└── ...

Reports are cached - use force_refresh: true to regenerate.

Environment Variables

Provider Fallback: If the preferred provider is unavailable, falls back to the other provider. If neither API key is set, uses rule-based analysis.

Related Skills

• pipeline-debug - General pipeline debugging
• cluster-quality - Quality metrics interpretation