retellai-advanced-troubleshooting
Apply advanced debugging for hard-to-diagnose issues including audio quality, latency, and transcription accuracy
Directory Structure
retellai-advanced-troubleshooting/
├── SKILL.md
└── examples/
└── example.py
File Descriptions
| File | Type | Purpose |
|---|---|---|
| SKILL.md | Markdown | Core skill instructions for advanced troubleshooting techniques |
| examples/example.py | Python | Example diagnostic scripts for deep investigation |
Summary
Category: advanced
Target Audience: Senior SREs, Staff engineers, Support escalation engineers
Trigger Phrases: retell hard bug, retell mystery error, retell audio issues, difficult retell issue
What This Skill Does
This skill provides advanced troubleshooting techniques for difficult Retell AI issues. It covers audio quality analysis using FFprobe and spectrum analysis, latency profiling across the voice pipeline, transcription accuracy investigation, WebSocket connection debugging, and creating minimal reproductions for support escalation.
Technical Success Criteria
- Comprehensive debug bundle created
- Failure layer identified (audio, STT, LLM, TTS, network)
- Minimal reproduction created
- Support escalation submitted with full context
- Resolution path identified
Business Success Criteria
- Resolved edge-case issues that block production
- Improved system understanding
- Successfully diagnose 95% of escalated issues
Related Skills
- retellai-debug-bundle - Initial diagnostic collection
- retellai-incident-runbook - Incident response procedures
- retellai-realtime-streaming - WebSocket debugging