sentry-code-review

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

• Third-party content exposure detected (high risk: 1.00). The skill's workflow explicitly fetches GitHub PR review comments via the GitHub API or WebFetch (gh api repos/{owner}/{repo}/pulls/.../comments) and parses the comment body — including the "🤖 Prompt for AI Agent" section — to guide analysis and edits, which exposes the agent to untrusted, user-generated third-party content that can alter its actions.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill fetches Sentry review comments at runtime from the GitHub PR endpoints (e.g. https://api.github.com/repos/{owner}/{repo}/pulls/<PR_NUMBER>/comments and/or the PR URL like https://github.com/{owner}/{repo}/pull/<PR_NUMBER>), and those comment bodies include an "🤖 Prompt for AI Agent" section whose text would directly control the agent's prompts — a required runtime dependency.