repomix

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly supports processing remote public repositories (e.g., SKILL.md "Remote Repository Support" with npx repomix --remote https://github.com/owner/repo and scripts/repomix_batch.py that accepts remote GitHub URLs), so it ingests untrusted, user-generated third‑party code and packages it for AI analysis — content that the agent will read and that can materially influence downstream analysis or actions, creating a risk of indirect prompt injection.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill invokes repomix at runtime (e.g., via "npx repomix --remote owner/repo" and the documented https://github.com/yamadashy/repomix), which causes remote code to be fetched/executed (npx/npm package) and remote repository contents to be pulled and packaged into LLM context — meeting the criteria for a runtime external dependency that can execute code and control prompts.