Langfuse Prompt and Trace Debugger
MANDATORY skill when KeyError or schema errors occur. Fetch actual prompt schemas instead of guessing. Use for debugging traces and understanding AI model behavior.
Install
/plugin marketplace add Freddy-CHEN-ux/00_claude_skillsThe repo is a Claude Code plugin marketplace. Run /plugin afterwards to pick the plugin that includes this skill.
npx skills add Freddy-CHEN-ux/00_claude_skills --skill langfuse-prompt-and-trace-debuggergit clone --depth 1 https://github.com/Freddy-CHEN-ux/00_claude_skills.git /tmp/00_claude_skills
mkdir -p ~/.claude/skills
cp -r /tmp/00_claude_skills/skills/langfuse-prompt-and-trace-debugger ~/.claude/skills/langfuse-prompt-and-trace-debuggerPersonal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.
From the SKILL.md
# Langfuse Prompt & Trace Debugger - MANDATORY FOR KEYERRORS ## 🔥 CRITICAL: Use This Skill Immediately When: **MANDATORY triggers (you MUST use this skill):** - ❗ Tests fail with `KeyError` (e.g., `KeyError: 'therapist_response'`) - ❗ Schema validation errors - ❗ Unexpected prompt response structure - ❗ Need to understand what fields a prompt actually returns **Common triggers:** - User asks to view a specific prompt - Code references a prompt but logic is unclear - Investigating why AI behavior doesn't match expectations - Debugging Langfuse traces - Analyzing model output in production **🚀 PROACTIVE TRIGGERS (non-obvious scenarios where you should automatically use this skill):** 1. **Production Debugging Queries:** - "Why didn't user X get a message?" - "Why did this intervention not fire?" - "What errors happened in production [timeframe]?" - "Debug this trace ID from Slack alert" - User reports: "AI didn't respond" or "got wrong response" - **→ Use `fetch_error_traces.py` to find error traces automatically** 2. **Performance & Cost Investigation:** - "Why are OpenAI costs high this week?" - "Which prompts are slowest?" - "Show me what happened during [time window]" - Job timeout errors in CloudWatch/logs - **→ Use `fetch_traces_by_time.py` to analyze patterns** 3. **Response Validation Issues:** - Logs show `_validation_error` metadata - "LLM returned unexpected structure" - Pydantic validation errors on AI responses
Details
- Skill name
langfuse-prompt-and-trace-debugger- Repository
- Freddy-CHEN-ux/00_claude_skills
- Path
skills/langfuse-prompt-and-trace-debugger/SKILL.md- Repo updated
- 10 January 2026
FAQ
What does the Langfuse Prompt and Trace Debugger skill do?
MANDATORY skill when KeyError or schema errors occur. Fetch actual prompt schemas instead of guessing. Use for debugging traces and understanding AI model behavior.
How do I install the Langfuse Prompt and Trace Debugger skill?
Claude Code plugin: /plugin marketplace add Freddy-CHEN-ux/00_claude_skills. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.
Is Langfuse Prompt and Trace Debugger popular?
It lives in Freddy-CHEN-ux/00_claude_skills, which has 8 GitHub stars and 3 forks.
Related skills
Add Integration
8Freddy-CHEN-ux/00_claude_skills
Build a new API integration for Nexus. Load when user mentions "add integration", "new integration", "integrate with", "connect to [service]", or "build [service] integration". Interactive workflow that discovers API endpoints, plans the integration, and creates a project for implementation.
AI Agents
8Freddy-CHEN-ux/00_claude_skills
Production-grade AI agent patterns with MCP integration, agentic RAG, handoff orchestration, multi-layer guardrails, and observability (modern best practices)
AI LLM
8Freddy-CHEN-ux/00_claude_skills
Production LLM engineering skill. Covers strategy selection (prompting vs RAG vs fine-tuning), dataset design, PEFT/LoRA, evaluation workflows, deployment handoff to inference serving, and lifecycle operations with cost/safety controls.
Langfuse Prompt Iterator
8Freddy-CHEN-ux/00_claude_skills
Comprehensive prompt iteration workflow for debugging and improving Langfuse prompts. Analyzes codebase usage, sets up reproducible tests, runs iterative refinement cycles with user feedback. Use when user says "prompt X isn't working right", "improve prompt", "iterate on prompt", "refine prompt", or "test prompt changes".
Incomplete Onboarding Assessor
8Freddy-CHEN-ux/00_claude_skills
Assess incomplete onboardings using COLD/WARM/HOT temperature model with time decay
Lighthouse Performance Optimization
8Freddy-CHEN-ux/00_claude_skills
Use when optimizing website performance. Run Google Lighthouse audits via MCP to measure metrics, identify bottlenecks, and iterate on improvements.
Browse the most popular Claude skills, all ai, llms & agents skills, or write your own with the skills generator.