Back home

Intel desk whatsmyagentdoing

whatsmyagentdoing

Original link

01

Intro

whatsmyagentdoing is an open-source (MIT) local agent dashboard: it shows the Claude Code, Codex and Gemini CLI sessions on your machine as live characters in a tiny isometric office — or as a plain list.\n\nGitHub repo Charonyuu/whatsmyagentdoing was created 2026-10-23 and has about 227 stars (v0.3.1). The design is deliberately minimal: it only reads the JSONL session logs those tools already write — no hooks, no API keys, and it does not intervene in the agents. It answers one question: what is every agent on this machine doing right now?

02

Explain

What it does\n\n- One screen, several tools: watch Claude Code, Codex and Gemini CLI sessions together\n- Two views: isometric little office (characters busy or done at their desks) or a plain list\n- Zero intrusion: reads existing local session logs; README stresses no hooks or API keys\n- Runs locally: data stays on your machine — useful when you have many agent terminals open and hate tab-hopping\n\n## Why it stands out\n\nAs local coding agents multiply, "who is stuck / who finished" becomes a real pain. whatsmyagentdoing pairs a memorable product name with a light, playful observability layer, and picked up early open-source attention (including Roundtable-style roundups).

03

Usage

Get started\n\n- Repo: https://github.com/Charonyuu/whatsmyagentdoing\n- Follow the README to install and launch the local dashboard; make sure Claude Code / Codex / Gemini CLI session-log dirs exist on the machine\n- Use the office view for a quick sweep of who is working or done; switch to the list for details\n\n## Who it's for\n\n- Developers running several local coding agents who want a one-glance overview\n- Anyone who wants read-only observability without changing agent configs\n\n> Still an early tool (~200+ stars); trust the README and current release for features.