add-adapter

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

• Third-party content exposure detected (medium risk: 0.30). SKILL.md’s workflow asks the user for a third-party GitHub repo URL and docs/README and then instructs the agent to read that text to generate catalog/docs content, so outsider-authored free text is ingested at runtime.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly asks for and requires reading the user's GitHub repository URL ("Their adapter's GitHub repository URL"), which would be fetched at runtime and whose repo/README content directly controls what the agent writes into the catalog/docs.