research-modes

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

• Third-party content exposure detected (high risk: 0.85). The skill’s workflow is explicitly about discovering and reading external sources (e.g., “Search for academic papers, technical documentation, blog posts… Check GitHub repositories, open-source projects, and code examples” and then “Read each source”), which at runtime can ingest outsider-authored free text from public web/GitHub pages into the LLM context during note-taking.