remote-project-control

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

• Third-party content exposure detected (high risk: 0.80). The skill's required workflow explicitly runs network commands and reads remote, user-controlled content (e.g., SKILL.md Step 3 and Step 4: ssh/remote-cmd/remote-bash to inspect server repo state and logs, and references/operations.md and templates which call out gh/GitHub API checks), so it ingests untrusted third‑party/user-generated content (server files, repo contents, GitHub API responses) and uses that data to decide subsequent actions.