gh

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

• Third-party content exposure detected (high risk: 1.00). The SKILL.md explicitly instructs use of gh commands that fetch and display GitHub content (e.g., "gh issue view", "gh issue list", "gh pr ...", "gh run view", and gh-sub-issue with repository URLs), which are untrusted, user-generated third-party pages/comments the agent would read and that could materially influence subsequent actions.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill instructs users to install the gh extension yahsan2/gh-sub-issue (https://github.com/yahsan2/gh-sub-issue) at runtime via "gh extension install", which fetches and installs remote code that the skill relies on for sub-issue functionality and thus executes external code.