telecode is a tool for running coding agents on your own machine, with an end-to-end-encrypted path between your browser and that machine. We take its security seriously and appreciate responsible disclosure.
Please do not report security vulnerabilities through public GitHub issues, discussions, or pull requests.
Instead, report privately through either channel:
- GitHub private vulnerability reporting (preferred): go to the repository's Security tab → Report a vulnerability. This opens a private advisory visible only to you and the maintainers.
- Email:
security@telecode.io.
Please include, as far as you can:
- the component affected (web, relay, daemon, protocol/crypto),
- a description of the issue and its impact,
- steps to reproduce or a proof of concept,
- any suggested remediation.
We will acknowledge your report within 5 business days, keep you updated on our assessment and a fix timeline, and credit you in the advisory once a fix ships (unless you prefer to remain anonymous).
Most relevant to telecode's trust model:
- Anything that lets the relay read plaintext session content (it must only ever see ciphertext plus routing metadata).
- Bypassing the approval gate so a tool call runs on a user's machine without a human decision.
- Authentication / authorization flaws in pairing, device tokens, channel tokens, or session access.
- Letting one user read or control another user's devices or sessions.
- Secret or key material exposure (in logs, storage, or over the wire).
Out of scope: issues that require a fully compromised user machine, social-engineering attacks, and
volumetric DoS against a relay you do not operate (self-host for full control — see
docs/self-hosting.md).
telecode is pre-1.0 and ships from main. Security fixes land on main and in the latest release; please
test against the latest before reporting.
The threat model documents exactly what each component can and cannot see, including the honest caveat about relay-visible routing metadata, and how to verify the relay only handles ciphertext.