triage

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

• Third-party content exposure detected (high risk: 1.00). The skill explicitly fetches and processes public, user-generated third-party content (see reproduce.md: "gh issue view" to load GitHub issue details, "git clone https://github.com/..." for GitHub repo URLs, "curl https://api.github.com/gists/..." for Gists, and "stackblitz-clone" for StackBlitz projects), and that content is required input the agent must read and act on as part of the triage workflow, which could enable 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 explicitly instructs cloning and running external GitHub repositories at runtime (e.g. "git clone https://github.com/<owner>/<repo>.git " followed by pnpm install/build), which fetches remote code that will be executed as part of the triage workflow, so https://github.com//.git is a runtime dependency that can execute remote code.