deepgit-search-guide

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

• Third-party content exposure detected (high risk: 0.80). This skill's workflow explicitly instructs searching and ingesting content from public third-party sources (e.g., GitHub, Papers with Code, arXiv, Semantic Scholar, author GitHub profiles and academic blogs/Twitter) as part of repository discovery and evaluation, meaning the agent would read and act on untrusted, user-generated web content.