AI Status

A local web dashboard where Claude Code (or any agent) reports its progress while it works. The agent writes Markdown to a session file; this app watches the file and renders it live in your browser — so you always know what your agent is doing, without babysitting the terminal.

Why you'd want it

• See progress at a glance. Each session is a Markdown file that re-renders live (SSE) every time the agent touches it, with changed blocks highlighted in amber.
• Run the agent right there. Open a Claude console inside each session's column — terminals live on the server and survive tab switches.
• No setup for the agent. The bundled status-orchestrator skill is embedded in the app and loaded automatically, so Claude maintains the status file and delegates work without you installing anything.
• Stays out of the way. System-tray icon, single instance, dark/light themes, desktop notifications, mobile drawer.

→ Full feature tour: docs/features.md

Install & run

1. Get it. Download the prebuilt binary for your OS from Releases, or build from source:
   • Windows: go build -ldflags="-H windowsgui" -o ai-status.exe
   • Linux / macOS: ./scripts/build.sh (or go build -o ai-status .)
2. Run it. ai-status.exe (Windows) or ./ai-status (Linux / macOS). It opens the dashboard at http://127.0.0.1:7879 and adds a tray icon.
3. Find it later. On Windows it auto-registers on first run, so "AI Status" shows up in the Start menu and Settings → Installed apps. On Linux run ./scripts/install-desktop.sh for an app-menu entry.

→ Full install details, runtime flags, and OS-specific notes: docs/install.md

Use it

1. Click New session — a fresh .md is created under sessions/.
2. Click the session title to copy its absolute path.
3. Paste that path into Claude and ask it to use it — the skill picks it up automatically.
4. Work as normal. Claude writes status; the page updates live.

→ Usage, the companion skill, and data layout: docs/usage.md

License

MIT