Skip to content

Security: trappsquid/tinycmdr-dev

Security

SECURITY.md

Security Policy

tinycmdr is an agent harness: it holds API tokens in .env, executes shell commands, edits files, and serves a token-gated web page. I take reports about these surfaces seriously.

Reporting a Vulnerability

Please do not open a public issue for a security report.

GitHub Private Vulnerability Reporting is the sole channel: the repository's Security → Advisories → Report a vulnerability button. If that button is not available to you, open a normal issue saying only "I have a security report" and I (@trappsquid) will open a private advisory to continue there - never post the details publicly.

Include: the version or commit you tested, the affected component, a description of the impact, and reproduction steps or a proof of concept. You will receive an acknowledgement within a few days; I aim to confirm or reject within two weeks and will coordinate disclosure timing with you.

Scope

Especially interested in:

  • Token and secret handling — anything that gets a token into a prompt, a log, a transcript, a chat message, or the web page
  • Path traversal / arbitrary file access through the file tools, uploads, downloads, or the chat lanes
  • Web page auth — token-gating bypasses, session fixation, cross-host exposure when bound to 0.0.0.0
  • Command injection through tool-call argument parsing and repair
  • Update flow — the SHA256SUMS verification is checksumming, not signing; reports about the integrity model are welcome
  • Installer privilege handling — anything that runs as root on Linux/macOS

Out of scope

  • Vulnerabilities in a self-hosted model you connect tinycmdr to
  • Issues requiring physical access to the host
  • The documented warning that a 0.0.0.0 bind sends the token in cleartext on the LAN — that is a stated, deliberate trade-off, not a bug

Supported Versions

Only the latest release receives security fixes. Update with tinycmdr update.

There aren't any published security advisories