fcode-code-validation

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

• Third-party content exposure detected (medium risk: 0.30). The skill clones and reads workspace code, configuration, and app metadata, which can contain outsider-authored free text if an attacker has committed malicious code or data to the repository.

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

• Potentially malicious external URL detected (medium risk: 0.30). The skill instructs the agent to install an external package via pnpm install -g @factorialco/fcode-cli during the acquisition procedure, but this registry transport does not provide strong first-party provenance for the third-party package name.