Skip to content

Security: seafresh71/juror

Security

SECURITY.md

Security policy

Supported versions

Security fixes are released on the latest 1.x version of Juror. Older releases and unreleased forks are not supported. Because the GitHub Action executes the source at the revision selected by each consumer, users should update their immutable pin after a security release rather than follow a moving tag.

Report a vulnerability privately

Do not open a public issue for a suspected vulnerability or include live credentials in a report. Use GitHub's private vulnerability reporting form. If the form is unavailable, contact a repository owner through GitHub without disclosing the technical details publicly and ask for a private reporting channel.

Include the affected version or commit, impact, reproduction steps, and any suggested mitigation. Use synthetic credentials and a private test repository whenever possible.

Response expectations

  • We aim to acknowledge a report within three business days.
  • We aim to provide an initial severity assessment within seven business days.
  • For an accepted report, we will share status at least every seven business days until a fix or documented mitigation is available.
  • We will coordinate disclosure and credit with the reporter. Fix timing depends on severity and release risk; these targets are communication commitments, not a guaranteed resolution deadline.

Scope

Reports are in scope when they affect Juror's CLI, GitHub Action, release artifacts, credential isolation, sandbox boundaries, review publishing permissions, or supply chain. Vulnerabilities in an upstream model provider or CLI should also be reported upstream; please report them here when Juror's integration materially increases the impact.

Prompt injection that can only influence review prose is an expected residual risk. Injection that exposes credentials, escapes a filesystem boundary, changes repository state, or obtains additional GitHub permissions is in scope.

See the threat model for trust boundaries, controls, and residual risks.

There aren't any published security advisories