Please report security issues privately, not in a public issue.

Use GitHub's private vulnerability reporting: the repository's Security tab → Report a vulnerability. That opens a private advisory visible only to the maintainers.

Please include: what you found, where (file and line if you can), how to reproduce it, and the impact you think it has. A working proof-of-concept helps but is not required.

You can expect an acknowledgement within a few days. Fixes for confirmed issues are prioritised over feature work.

This project is a transparent, self-hostable operations dashboard. Its design splits public reads from local-only writes/control. The security model that matters most:

• The public surface is intended to be read-only plus a small set of clearly public write endpoints (POST /api/signals, POST /api/signatures/sign). Everything else — terminal, chat dispatch, kill switch, approvals, handoffs — must be reachable only by a local caller.
• The trust boundary is the fleet front door (fleet/bin/fleet.py): it decides local vs remote and sanitises X-Forwarded-For before forwarding to the cockpit. A remote caller must never reach a control or local-only action.

Findings that cross that boundary — a remote caller reaching a control path, spoofing local identity, or reading private data from a public endpoint — are the highest priority.

This project is pre-1.0 and moves fast. Security fixes are applied to the main branch and the latest tagged release. There is no long-term support branch yet.

If you self-host and expose it to the internet (e.g. via a tunnel), keep the host patched, do not run it on a machine whose compromise you cannot afford, and treat the "writes are local-only" boundary as load-bearing — it is the only thing between the public and your terminal.