TraceLens
Runtime truth engine for web applications, causality over metrics, AI-queryable.
What is TraceLens?
TraceLens transforms application debugging by focusing on causality rather than metrics. Instead of telling you "your app is slow," it builds real-time dependency graphs from browser and server runtime signals to show you exactly why it's slow.
It is self-hosted (your data never leaves your infrastructure), open-source under MIT, and integrates with AI coding assistants through the Model Context Protocol, so you can ask Claude or Kiro questions like "what are my app's current performance bottlenecks?" and get precise, grounded answers.
The most interesting design decision: TraceLens monitors its own dashboard performance in real time, so you experience the platform's capabilities before writing a single line of integration code.
Who is it for?
- Developers using AI coding assistants (Claude, Kiro, Cursor, Windsurf) who want grounded performance context
- Vibe-coding teams that don't want to wire up DataDog / New Relic just to see what's slow
- Teams who care about causality (why is this slow?) more than dashboards (here are 50 metrics)
- Anyone shipping browser apps where data privacy matters, TraceLens stays self-hosted
What does it do?
- Real-time dependency graph from browser + server runtime signals
- Critical-path detection (which exact call is blocking the response)
- Web Vitals + automatic performance instrumentation out of the box
- AI-queryable via MCP server, natural-language perf questions
- Self-monitoring meta-observability (the dashboard monitors itself)
- 5+ install modes: standard, quick, demo, enhanced, custom-port
- Localhost-first deployment, no cloud dependency
How is it built?
What makes it interesting?
- Builds real causality graphs, not just metric dashboards
- < 1 ms browser SDK overhead, you don't pay for the observability
- Reduces AI debugging cost by ~80% via precise MCP context
- Self-monitors the dashboard so you see it working before integrating
- MIT licensed and fully self-hostable
Want something like this for your business?
We adapt the patterns above to your stack on every engagement. The 15-minute discovery call is free, you leave with a plan regardless.