antechamber is a local development tool. It runs a coding agent (pi or OmO) that can execute arbitrary shell commands and edit files on the machine it runs on, on behalf of whatever OpenChamber instance is talking to it. Treat it like a local dev server with shell access, not a hardened network service.
Never expose antechamber to an untrusted network without authentication.
By default it binds to loopback (127.0.0.1) with no auth, which is safe for
local use. Binding to a non-loopback address is expected to require a
password or an explicit opt-out flag; see docs/configuration.md
and the README's Security section for the current state of that work, which
is still landing.
If you run antechamber behind a reverse proxy, on a shared machine, or on a network reachable by other users, you are responsible for authentication, TLS termination, and network-level access control in front of it.
antechamber is pre-1.0. Only the latest published release and the main
branch receive security fixes. There is no long-term support branch yet.
| Version | Supported |
|---|---|
| latest release | yes |
| older releases | no |
Please report security issues privately through GitHub's private vulnerability reporting for this repository, rather than opening a public issue.
Include:
- A description of the issue and its impact
- Steps to reproduce (a minimal repro is ideal, since this is a local tool)
- The antechamber version and backend (pi or OmO) involved
We'll acknowledge reports as soon as we can and work with you on a fix and disclosure timeline before any public write-up.
- Vulnerabilities in pi, OmO, or OpenChamber themselves. Report those upstream.
- The fact that antechamber's agent backend can run shell commands. That's the product; the mitigation is network isolation and, once it lands, authentication, not sandboxing the agent itself.