gh-stack

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). Runtime workflow relies on the user/agent selecting and inspecting GitHub stack/PR state via gh stack view --json and related GitHub-targeted commands, which necessarily ingest outsider-authored free text from PRs (titles/bodies/comments included in PR state) returned by the GitHub API.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill directs runtime installation and use of the external GitHub CLI extension "github/gh-stack" (via the command gh extension install github/gh-stack), which fetches and installs remote executable code that the skill requires.