# agentacct

A local-first dashboard that shows what coding agents such as Claude Code, Codex and OpenCode actually did on a task, breaking each one into tool calls, files changed, tests run, time spent and estimated cost.

agentacct reads the local session logs of coding agents including Claude Code, Codex, Kimi Code and OpenCode, and turns them into a browsable record of recent activity across a person's projects and clients. Opening a task shows a Work Receipt: an activity timeline from the agent's first attempt to its result, with the tools used, files changed, tests run, elapsed time and estimated token cost, and full check details one click away.

It distinguishes what an agent reported as done from what was actually verified by passing checks after the last recorded change, and keeps a failed attempt visible in history alongside any later, successful result rather than only showing the latest state. A work group view aggregates every agent's activity in a shared project folder into one timeline, and a dashboard surfaces the most recently active tasks across all tracked projects.

It installs with pipx, requires no account, and sends nothing off the machine; a signed macOS app is also available for people without Python.

Website: https://trytofu.ai