cloud-weaver-teardown

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

• Third-party content exposure detected (medium risk: 0.30). The required workflow interacts with GitHub repositories via gh CLI, including listing workflows, triggering runs, and watching logs, which can expose the agent to outsider-authored free text from issues, PRs, or workflow steps.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill instructs the agent to download an arbitrary script from https://cloudweaver.fagnerlopes.dev/teardown.py during a fallback scenario, introducing a weak/unknown/suspicious runtime external dependency.