Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Collect Langfuse debug evidence for support tickets and troubleshooting. Use when encountering persistent issues, preparing support tickets, or collecting diagnostic information for Langfuse problems. Trigger with phrases like "langfuse debug", "langfuse support bundle", "collect langfuse logs", "langfuse diagnostic", "langfuse troubleshoot".
.claude/skills/jeremylongshore-langfuse-debug-bundle/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 37% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 49% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 220% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 89% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 26% | 0% |
!node --version 2>/dev/null || echo 'N/A' !python3 --version 2>/dev/null || echo 'N/A' !npm list langfuse @langfuse/client @langfuse/tracing 2>/dev/null | head -5 || echo 'No langfuse packages'
Collect all diagnostic information needed for Langfuse support tickets: environment versions, SDK config, API connectivity, redacted logs, and a reproduction template.
Save this as langfuse-debug.sh and run it:
bash#!/bin/bash set -euo pipefail BUNDLE_DIR="langfuse-debug-$(date +%Y%m%d-%H%M%S)" mkdir -p "$BUNDLE_DIR" echo "=== Langfuse Debug Bundle ===" | tee "$BUNDLE_DIR/summary.txt" echo "Generated: $(date)" | tee -a "$BUNDLE_DIR/summary.txt" # --- Environment --- { echo "" echo "--- Environment ---" echo "Node.js: $(node --version 2>/dev/null || echo 'not installed')" echo "Python: $(python3 --version 2>/dev/null || echo 'not installed')" echo "npm: $(npm --version 2>/dev/null || echo 'not installed')" echo "OS: $(uname -srm)" } >> "$BUNDLE_DIR/summary.txt" # --- SDK Versions --- { echo "" echo "--- SDK Versions ---" npm list langfuse @langfuse/client @langfuse/tracing @langfuse/otel @langfuse/openai @langfuse/langchain 2>/dev/null || echo "npm: no langfuse packages" pip show langfuse 2>/dev/null | grep -E "Name|Version" || echo "pip: langfuse not found" } >> "$BUNDLE_DIR/summary.txt" # --- Config (redacted) --- { echo "" echo "--- Langfuse Config ---" echo "LANGFUSE_PUBLIC_KEY: ${LANGFUSE_PUBLIC_KEY:+SET (${LANGFUSE_PUBLIC_KEY:0:12}...)}" echo "LANGFUSE_SECRET_KEY: ${LANGFUSE_SECRET_KEY:+SET}" echo "LANGFUSE_BASE_URL: ${LANGFUSE_BASE_URL:-NOT SET}" echo "LANGFUSE_HOST: ${LANGFUSE_HOST:-NOT SET}" } >> "$BUNDLE_DIR/summary.txt" # --- Network Connectivity --- { echo "" echo "--- Network Test ---" HOST="${LANGFUSE_BASE_URL:-${LANGFUSE_HOST:-https://cloud.langfuse.com}}" echo "Target host: $HOST" echo -n "Health endpoint: " curl -s -o /dev/null -w "%{http_code} (%{time_total}s)" "$HOST/api/public/health" 2>/dev/null || echo "FAILED" echo "" if [ -n "${LANGFUSE_PUBLIC_KEY:-}" ] && [ -n "${LANGFUSE_SECRET_KEY:-}" ]; then AUTH=$(echo -n "$LANGFUSE_PUBLIC_KEY:$LANGFUSE_SECRET_KEY" | base64) echo -n "Auth test: " curl -s -o /dev/null -w "%{http_code}" \ -H "Authorization: Basic $AUTH" \ "$HOST/api/public/traces?limit=1" 2>/dev/null || echo "FAILED" echo "" fi } >> "$BUNDLE_DIR/summary.txt" # --- Application Logs (redacted) --- if [ -d "logs" ]; then grep -i "langfuse\|trace\|generation\|flush" logs/*.log 2>/dev/null | \ tail -100 | \ sed 's/sk-lf-[a-zA-Z0-9]*/sk-lf-***REDACTED***/g' | \ sed 's/pk-lf-[a-zA-Z0-9]*/pk-lf-***REDACTED***/g' \ > "$BUNDLE_DIR/app-logs-redacted.txt" 2>/dev/null || true fi # --- Package Dependencies --- if [ -f "package.json" ]; then grep -A 100 '"dependencies"' package.json | head -60 > "$BUNDLE_DIR/package-deps.txt" 2>/dev/null || true fi # --- Reproduction Template --- cat > "$BUNDLE_DIR/reproduction-steps.md" << 'REPRO' # Bug Report ## Environment - Node.js version: - Langfuse SDK version: - Langfuse host: Cloud / Self-hosted (version: ) ## Steps to Reproduce 1. 2. 3. ## Expected Behavior ## Actual Behavior ## Error Messages
Paste error output here
## Relevant Code// Paste minimal reproduction here
REPRO
# --- Package ---
tar -czf "$BUNDLE_DIR.tar.gz" "$BUNDLE_DIR" 2>/dev/null
echo ""
echo "Bundle created: $BUNDLE_DIR.tar.gz"
echo "Contents:"
ls -la "$BUNDLE_DIR/"
Always redact before submitting:
| Include | Redact | |---------|--------| | Error messages and stack traces | API keys (pk-lf-*, sk-lf-*) | | SDK versions and config structure | Secret keys and passwords | | HTTP status codes | PII (emails, user IDs) | | Timing and latency data | Internal URLs and IPs | | OS and runtime versions | Database connection strings |
bash langfuse-debug.shreproduction-steps.mdFor a fast check without creating files:
bashset -euo pipefail echo "=== Quick Langfuse Check ===" echo "Node: $(node --version 2>/dev/null || echo N/A)" npm list langfuse @langfuse/client 2>/dev/null | head -5 echo "" echo "Public Key: ${LANGFUSE_PUBLIC_KEY:+SET} | Secret Key: ${LANGFUSE_SECRET_KEY:+SET}" HOST="${LANGFUSE_BASE_URL:-${LANGFUSE_HOST:-https://cloud.langfuse.com}}" echo "Health: $(curl -s -o /dev/null -w '%{http_code}' $HOST/api/public/health)"
| Collected Item | Why It Matters | |----------------|---------------| | SDK version | Version-specific bugs, breaking changes between v3/v4/v5 | | Environment versions | Node 18+ required, Python 3.9+ | | Network test | Firewall, DNS, self-hosted connectivity | | Auth test | Key validity, project mismatch | | Redacted logs | Trace errors, flush failures, rate limits | | Package deps | Conflicting versions, missing peer deps |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 25,946 | 25,119 | -3% | 1 | 1 | 0% | 4,195 | 5,766 | +37% | 0 | 0 | — |
case-02 | fail→pass | 15,629 | 12,050 | -23% | 1 | 1 | 0% | 2,204 | 3,277 | +49% | 0 | 0 | — |
case-03 | fail→pass | 54,725 | 22,641 | -59% | 1 | 1 | 0% | 1,728 | 5,535 | +220% | 0 | 0 | — |
case-04 | fail→fail | 41,068 | 16,515 | -60% | 1 | 1 | 0% | 1,189 | 3,708 | +212% | 0 | 0 | — |
case-05 | fail→pass | 33,915 | 10,270 | -70% | 1 | 1 | 0% | 1,476 | 2,792 | +89% | 0 | 0 | — |
case-06 | fail→fail | 11,051 | 7,955 | -28% | 1 | 1 | 0% | 1,060 | 2,242 | +112% | 0 | 0 | — |
case-07 | pass→pass | 12,610 | 7,778 | -38% | 1 | 1 | 0% | 1,179 | 2,362 | +100% | 0 | 0 | — |
case-08 | pass→pass | 23,122 | 15,694 | -32% | 1 | 1 | 0% | 3,444 | 3,840 | +11% | 0 | 0 | — |
case-09 | pass→pass | 19,548 | 10,105 | -48% | 1 | 1 | 0% | 2,288 | 2,646 | +16% | 0 | 0 | — |
case-10 | fail→pass | 18,281 | 3,233 | -82% | 1 | 1 | 0% | 1,824 | 2,304 | +26% | 0 | 0 | — |
case-11 | fail→pass | 15,066 | 7,837 | -48% | 1 | 1 | 0% | 1,650 | 2,315 | +40% | 0 | 0 | — |
case-12 | pass→pass | 8,047 | 8,589 | +7% | 1 | 1 | 0% | 511 | 2,478 | +385% | 0 | 0 | — |
case-13 | fail→pass | 15,963 | 10,331 | -35% | 1 | 1 | 0% | 1,722 | 2,592 | +51% | 0 | 0 | — |
case-14 | fail→pass | 12,066 | 6,100 | -49% | 1 | 1 | 0% | 1,746 | 2,736 | +57% | 0 | 0 | — |
case-15 | fail→pass | 14,656 | 4,961 | -66% | 1 | 1 | 0% | 1,581 | 2,304 | +46% | 0 | 0 | — |
case-16 | pass→pass | 10,847 | 7,999 | -26% | 1 | 1 | 0% | 903 | 2,359 | +161% | 0 | 0 | — |
case-17 | fail→pass | 20,281 | 14,021 | -31% | 1 | 1 | 0% | 1,988 | 3,007 | +51% | 0 | 0 | — |
case-18 | fail→pass | 21,565 | 2,192 | -90% | 1 | 1 | 0% | 2,201 | 2,237 | +2% | 0 | 0 | — |
case-19 | fail→pass | 23,923 | 2,757 | -88% | 1 | 1 | 0% | 3,072 | 2,343 | -24% | 0 | 0 | — |
case-20 | fail→pass | 8,361 | 11,323 | +35% | 1 | 1 | 0% | 1,559 | 3,187 | +104% | 0 | 0 | — |
case-21 | pass→pass | 7,422 | 9,199 | +24% | 1 | 1 | 0% | 1,186 | 2,341 | +97% | 0 | 0 | — |
case-22 | pass→pass | 18,471 | 5,341 | -71% | 1 | 1 | 0% | 2,028 | 2,843 | +40% | 0 | 0 | — |
case-23 | pass→pass | 19,299 | 20,548 | +6% | 1 | 1 | 0% | 3,524 | 4,869 | +38% | 0 | 0 | — |
case-24 | pass→pass | 23,271 | 17,579 | -24% | 1 | 1 | 0% | 2,879 | 4,397 | +53% | 0 | 0 | — |
case-25 | pass→pass | 16,360 | 14,367 | -12% | 1 | 1 | 0% | 1,757 | 3,639 | +107% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 25 cases were attempted, and 22 counted toward the lift figure. The other 3 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +52 percentage points is the difference between those two pass rates over the 22 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.