Back to directory

agenttrail

agenttrail is a local observability tool for AI coding agents that watches project files and agent activity and turns them into two live views: a project structure map and a 3D task-and-role visualization.

agenttrail

agenttrail gives visibility into what AI coding agents such as Codex and Claude Code are doing, without piecing together several terminal windows. It watches local files, plans and supported agent events, then visualizes that evidence; it does not run the agents, assign their work, or decide when a task is finished.

The project has two views. Agenttrail Map shows project structure, progress and which components are changing, sourced from a PLAN.md file, file changes and optional Claude Code hooks, started with npx agenttrail. Agenttrail Kitchen visualizes current tasks, role contributions and completed work as a 3D scene of chefs, order tickets and deliveries, sourced from native agent todos and optional hooks, started with npx agenttrail-kitchen.

Each view runs independently in the browser with its own local service, requires Node.js 20+ and a WebGL-capable browser, and needs no API key for basic use.

More in AI evals

See all 269
  • Provenance gate

    github.com

    A deterministic authorization gateway that sits between an AI agent and its tools, tracing where each parameter of a side-effect call came from, and blocking the call or asking a human when the source does not check out.

  • JevBench

    benchmarkheaven.com

    A reproducible benchmark from Benchmark Heaven that ranks Jev-class typed decision models on intelligence, calibration, speed, and cost, with a published harness, scoring rules, and results.

  • A co-evolutionary adversarial red-teaming tool for AI agents that uses quality-diversity evolutionary search to autonomously breed edge-case scenarios, testing whether an agent with tool access can be made to hallucinate, deadlock or leak funds.