forge-implement-issue

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

• Third-party content exposure detected (high risk: 1.00). The skill's Step 1 explicitly instructs the agent to run "gh issue view <ISSUE_NUMBER>" and parse the GitHub issue title, body, labels, and comments (user-generated, public third‑party content), which the agent is expected to read and use to plan and drive implementation decisions.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill calls out to GitHub at runtime (e.g., fetching an issue via "gh issue view" or an issue URL like https://github.com/owner/repo/issues/123) and uses the issue body/requirements to drive implementation decisions, so remote content can directly control the agent's instructions.