skep

Govern your AI coding agent. Sandbox it, verify its work, approve before it lands.

skep is a local-first supervisor for AI coding agents — Claude Code, Codex, Aider, or its own built-in worker. Every agent works behind a contract: sandboxed execution in a disposable worktree, independent re-verification of whatever the agent claims, and a patch that reaches your repository only when you approve it. It is v1, single-operator by design, and has been daily-driven in the field since 2026-07-17.

Install

pipx install skep

Or uvx skep to try it without installing. On Debian/Ubuntu and other PEP 668 distros, the system Python refuses bare pip install with an externally-managed-environment error — install skep as an app instead: sudo apt install pipx && pipx install skep, then pipx ensurepath. Inside a virtual environment, pip install skep works as ever. Please don't reach for --break-system-packages — least privilege applies to installers too. From a checkout, bash scripts/install.sh detects the source tree and runs uv sync. Linux sandboxing needs bubblewrap (sudo dnf install bubblewrap / sudo apt install bubblewrap); macOS uses Seatbelt out of the box.

60-second quickstart

skep serve   # daemon + web UI at http://127.0.0.1:8765
skep chat    # the same conversation in your terminal

Tell the chat what repo to work on and what you want done — approvals, cards, and live run events arrive inline. Or skip the conversation:

skep run /path/to/your-repo "fix the failing test" --execution-mode workspace
skep review <task_id>             # inspect the patch + evidence
skep review <task_id> --approve   # lands on branch skep/<task_id>, never main

The complete first-run path, worker setup, and sandbox notes are in docs/quickstart.md.

The security model

The agent never touches your repository directly. It works in an isolated git worktree, produces a patch plus evidence, and the patch lands only through your approval — landing IS the commit.

• Sandbox. macOS Seatbelt or Linux bubblewrap confines every worker; filesystem and network access are policy, not habit.
• Verify. skep re-runs the project's pinned verify command on a clean copy before you review — the worker's own claim is never the evidence.
• Approve. A patch, not a promise. Workers cannot push, pull, fetch, or switch branches; no permission grant overrides that.
• Audit. Every run, event, shell command, approval, and decision is a durable record you can inspect later.

When a worker needs more permission than policy grants, skep stops at an approval gate: approve once, deny, or approve-and-remember so similar future runs move with less friction while the ledger still records everything (docs/approvals.md).

Status

The command deck

The chat composer doubles as a deterministic command line: any message starting with / is parsed by the UI and executed against the same HTTP API the buttons use — the assistant's model never sees it. /help lists the deck; reads (/policy, /state, /runs, /approvals, /repos) render immediately, and mutations (/setup, /workon, /phase, /land, /approve, /deny, /schedule) show the same confirmation card a model proposal would, audited under actor operator-command. /workon <path> makes any local directory a first-class workspace: a confirmed git init + baseline commit if needed, then the same trusted-project setup a registered repo gets. The same deck works in the terminal: skep chat (see docs/cli-reference.md).

Messenger channels

Operate skep from Telegram, Slack, or Discord: enable a channel in Settings, allow-list your chat id, paste the bot token, and an allow-listed message runs the same Queen turn — with the same approval gates — as the web composer. When a channel is confirm-enabled, low-risk actions resolve inline; shell commands, policy changes, and landings are never confirmable from a messenger. See docs/how-it-works.md.

Documentation

The curated index lives at docs/README.md — start with docs/quickstart.md. Worker adapters and the custom worker contract are in docs/workers.md; messenger channels in docs/how-it-works.md; sandbox backends in docs/sandboxing.md. Decisions are recorded in docs/adr/; the post-launch roadmap in docs/post-launch.md; development and PR guide in CONTRIBUTING.md; vulnerability reports via SECURITY.md.

Website: skep.anmolnoor.com

License

MIT.