retro

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

• Third-party content exposure detected (high risk: 0.75). The /retro skill’s runtime workflow ingests outsider-authored free text via git log/shortlog/commit-message parsing (e.g., PR titles and commit subjects from other contributors) into the LLM context for narrative generation; commit messages are not authored by the operating user and are treated as outsider text.