If you discover a security vulnerability in OpenChamber, please report it responsibly.

Email: security@openchamber.dev

Please include:

• Description of the vulnerability
• Steps to reproduce
• Affected version(s)
• Potential impact

I'll acknowledge receipt within 48 hours and aim to provide a fix or mitigation as quickly as possible.

Please do not open public GitHub issues for security vulnerabilities.

OpenChamber handles sensitive context including:

• UI authentication (password-protected sessions, JWT tokens)
• Cloudflare tunnel access (remote connectivity)
• Terminal access (PTY sessions)
• Git credentials and SSH keys
• File system operations

Security reports related to any of these areas are especially appreciated.

Signing in to OpenChamber gives you the same access as the user running it: a terminal, an agent with a shell, and file access anywhere that user can read or write. Checks inside a signed-in session, such as the workspace limits on file routes or how symlinks resolve, keep the interface predictable. They are not a security boundary, and bypassing them is not a vulnerability.

What I treat as a vulnerability:

• reaching any of that without signing in: authentication bypass, exposure on the network, tunnel, relay or pairing flaws, requests from other websites acting with your session;
• content you didn't write acting on your behalf: a repository, web page, agent output, HTML preview or third-party extension that runs code, reads files or credentials, or escapes its sandbox;
• credentials or conversation content leaving the machine somewhere you didn't send them.

Security fixes are applied to the latest release. There is no LTS or backport policy at this time.