acquiring-skills

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

• Third-party content exposure detected (medium risk: 0.30). The workflow requires searching, inspecting, and installing skills from community repositories, GitHub, and registries like ClawHub, which ingest outsider-authored free text into the agent's environment.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill describes an installation assistant that fetches instructions and community packages from third-party hubs like ClawHub and arbitrary GitHub repositories without enforced cryptographic verification, constituting a weak/unknown runtime dependency.