Agent AFK
Start a task and walk away. AFK builds the feature, verifies its own work, and texts you when it's done. You come back to a receipt of every step, every decision, and what's waiting on you.
Open source. Runs locally. Works with any model provider, including local models. Your keys, your machine, your rules.
<p align="center" <video src="https://github.com/user-attachments/assets/4a84f6e4-2494-4065-a01f-54a1c557f7f3" width="820" autoplay loop muted playsinline Your browser does not support the video tag. </video </p
⭐ Like the idea of an agent you can walk away from? Star the repo -- it's the fastest way to help other people find it.
Quick Start
npm install -g agent-afk # Node ≥ 22 required
afk login # authenticate (auto-detects Claude Code / Codex creds)
afk doctor # verify everything works
afk chat "hello" # first real conversationThat's it. You're in.
Try without installing: npx agent-afk chat "hello" runs a one-shot turn with zero global install.
Already using Claude Code or Codex? afk login will detect your existing credentials automatically. Run afk migrate to import your plugins, skills, and MCP servers too -- it live-reads the source tool's dirs, so anything you install there keeps showing up in AFK with no re-run.
📖 Full documentation at docs.agentafk.com -- quickstart, configuration, model setup, surfaces, skills, and SDK reference.
What you can do with it
- Chat from your terminal — afk chat "..." for one-shot, afk for a REPL with full tool access (Bash, file ops, web fetch, grep/glob, subagents).
- Hand long work off to a daemon — afk daemon runs headless. Pair it with sendtelegram and you get pings on your phone when work lands in a terminal state.
- Read the trace — every run writes an append-only record of what the agent did. afk trace show prints it back as a human-readable receipt — tool calls, gate decisions, subagent lifecycles, cost — so you can audit a run without reaching for jq.
- Message Claude from Telegram — afk telegram setup walks you through bot token + allowlist. After that you have a private chat surface backed by the same session manager as the REPL.
- Built-in orchestrators — /mint, /diagnose, /spec, /research, /ship, /review dispatch subagent waves. /mint takes a feature idea and runs spec → research → plan → parallelize → build → verify → ship. /diagnose forks parallel root-cause hypotheses for failing tests and bugs.
- Cross-session memory — Claude remembers preferences, decisions, and procedures across runs. See Memory below.
- Background subagent jobs — dispatch a subagent with mode:'background'; results auto-deliver into the model's context when they finish. /bgsub lists running and completed jobs, /bgsub:join <id replays a result manually.
How it compares
| | Claude Code | Agent AFK | |---|---|---| | Harness | Closed binary | Apache-2.0, editable | | The loop | You configure around it | You edit the loop | | Behavior | Mostly fixed | Prompts, gates, routing, skills are code | | Result | A great default agent | An agent system you own |
The model isn't the product -- the loop is. Agent AFK hands you the loop as code: prompts, gates, routing, skills, traces, providers, terminal states. Edit any of them.