prd-to-issues

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

• Third-party content exposure detected (high risk: 0.75). The skill’s runtime path explicitly fetches the PRD from a user-supplied GitHub issue/URL via gh issue view <number> (with comments), and that PRD content/comments are outsider-authored unless the operating user authored them, so free text from GitHub issue bodies/comments can be ingested into the agent’s LLM context.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill explicitly instructs fetching the PRD at runtime with "gh issue view " (or a GitHub issue URL like https://github.com/.../issues/), and the fetched PRD text is used to drive the agent's prompt and outputs, so this is a runtime external dependency that controls the agent's instructions.