shared-docs

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly fetches and updates git submodules from remote repositories (remote_url in .gitmodules and actions like git submodule update --remote in references/sync.md, git fetch/push in references/init.md and references/write.md), which are arbitrary user-generated third-party repos (GitHub/GitLab) whose contents the agent reads and uses to decide flows (e.g., diff-based small/large edit classification, PR creation, doctor checks), so untrusted third-party content can materially influence actions.