markdown-fixer

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

• Third-party content exposure detected (high risk: 0.70). The required workflow ingests a “structured scan report from markdown-scanner” and an “approved issue list” as input, which at runtime can include outsider-authored free text from issues/files that markdown-scanner scanned/classified (e.g., issue/PR/wiki/discussion body text), and that text is then read by the agent to generate fixes.