developer

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly fetches and processes public, user-generated content (e.g., remote_fetch.go functions like downloadIncludeFromWorkflowSpec / downloadFileFromGitHub, the Go Fan workflow that "research[es] GitHub Repo" to generate scratchpad/mods summaries, and validators that query PyPI/npm/docker registries), and that fetched third‑party content is parsed and used in compilation/validation flows—meaning untrusted external content can materially influence agent decisions.