Agents fail silently. You fire one off, it runs, nothing comes back - no trace, no cost data, no idea which call broke.
Tracea captures every tool call, LLM response, and cost spike. Automatic RCA tells you exactly why it failed. YAML detection rules catch loops, spikes, and silent errors before they hit production.
Self-hosted. One Docker command. No data leaves your network.
Company Brain turns every session into team memory - agents start smarter each run.
I ve been working on a project called ObservAgent and wanted to share it here to get feedback from people building with AI agents, especially Claude Code users.
ObservAgent is a local-first observability dashboard for Claude Code. It hooks into Claude's built-in hook system, zero code changes and streams every tool call, cost, latency, and subagent event to a live dashboard in real time.
Stop flying blind on expensive agent runs. See exactly what's happening, what it costs, and what's slow.
npm i -g @darshannere/observagent && observagent init
Tracea