build-prd

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

• Third-party content exposure detected (high risk: 0.90). The workflow explicitly instructs the agent to read and interpret GitHub issue bodies and repository files (via "gh issue view" and "cat project-instructions.md"/"README.md"), which are user-generated/untrusted content and are used to drive PRD decisions and subsequent GitHub actions, enabling indirect prompt injection.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill calls the GitHub CLI at runtime (e.g., "gh issue view ISSUE_NUMBER --repo ORG/REPO") to fetch issue/repository content that is then injected into the PRD-generation flow, so the external GitHub repo/issue data directly control prompts and are a required dependency.