opencut

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

• Third-party content exposure detected (high risk: 0.85). Yes: the required workflow step “Re-fetch README/CONTRIBUTING via scripts/fetch-docs.sh” performs runtime network fetches of outsider-authored GitHub README/.github/CONTRIBUTING text (public web content) and that fetched prose can be ingested into the agent/LLM context.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill's runtime script scripts/fetch-docs.sh fetches upstream README/CONTRIBUTING files from https://raw.githubusercontent.com/${OWNER_REPO}/main (via curl) and the skill instructs re-fetching those docs at runtime to inform responses, so this is a runtime fetch that can directly influence prompts.