evor

The Evor host agent — runs on your laptop, connects to the Evor platform, and executes the work that your AI sessions request: reading your repo, applying proposed diffs, committing under your git identity, pushing to GitHub.

Evor is conversation-first pair programming. You chat, AI plans + proposes changes, you approve, and the changes land as real commits on a branch on your machine. The agent is what makes the local half of that flow work. Without it, you get REST-only workflows (chat + PRs opened through the GitHub API, no local checkout, no live file updates).

Backend + web: see github.com/rabin-a/evor2.

Install

npm install -g @evorlive/evor
# or
pnpm add -g @evorlive/evor

After install, the evor command is on your PATH (the scope is only in the package name; the binary is just evor). Requires Node.js 20+.

Quick start

# 1. Point the agent at your Evor server
evor config set backend b.evor.live       # -> wss://b.evor.live/agent
# for local dev, default is ws://localhost:4000/agent

# 2. Link your personal pairing token (copy from /settings/agent in the UI)
evor link evor_pat_xxxxxxxxxxxxxxxxxxxx

# 3. Start the agent (background daemon)
evor start

# 4. Confirm it's online
evor status

Now any AI session you run on Evor will route its work to your local machine.

Commands

evor start              Start the agent in the background (daemon)
evor start --foreground Run in foreground (Ctrl+C to stop)
evor stop               Stop the daemon (also sweeps orphans)
evor restart            Stop, then start
evor status             Show pid, uptime, backend URL, token, paths
evor logs [-f] [-n N]   Tail the daemon log; -f follows; -n lines (80)
evor link <token>       Save your pairing token
evor unlink             Forget the saved token
evor token              Print the current token
evor config list        Show current config
evor config set <k> <v> Persist a config key (backend / name / workspaceRoot)
evor config unset <k>   Reset a key to its default

State

Everything the agent writes lives under ~/.config/evor/:

~/.config/evor/
├── token         your personal pairing token
├── config.json   { backend, name, workspaceRoot }
├── agent.pid     daemon process id
├── agent.json    { startedAt, backendUrl, agentName }
└── agent.log     daemon stdout + stderr

Cloned repositories land under ~/evor-workspaces/<owner>/<repo> by default (override with evor config set workspaceRoot /some/path).

How it works

The agent opens a persistent WebSocket to your Evor backend. When an AI session produces work, the backend routes it to your agent (matched by pairing token). The agent can:

• Run a CLI AI provider — claude -p … or codex exec … on your host, under your subscription / keychain auth
• Clone & fetch repos — git clone over HTTPS using your GitHub token
• Apply diffs to the working tree — files land on disk as the AI proposes them
• Run git operations — git add / commit / push / checkout-file against the task's branch
• Execute shell commands — only after the user approves, and only from an allowlist

Every destructive action requires explicit approval from you on the web UI. Chat history is private to the user that owns the session.

Security

• Each user has their own agent token. Work from user A never runs on user B's agent.
• Commands go through an allowlist (node, pnpm, npm, yarn, bun, go, python, git, ls, cat, grep, rg, make, bash, sh by default; override with CMD_ALLOWLIST).
• File writes are path-escape checked — nothing outside WORKSPACE_ROOT is touched.
• Git commits use your local ~/.gitconfig identity.

License

MIT