issue-triage

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

• Third-party content exposure detected (high risk: 1.00). This skill explicitly fetches and ingests user-generated GitHub issue bodies and comments (see Phase 1 data gathering: gh issue list --json ... body and Phase 2/Phase 2 fetch gh issue view {num} --json comments and the subagent prompt which includes "Body" and "Comments"), so untrusted third-party content is read and directly used to decide and execute triage actions.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill issues runtime GitHub API calls (e.g., via the gh CLI commands "gh issue list --json …" and "gh issue view {num} --json comments", which hit https://api.github.com/repos/{owner}/{repo}/issues and related endpoints) and then injects fetched issue bodies/comments directly into agent/subagent prompts, so external GitHub content can directly control the model's prompts at runtime.