secure

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

• Third-party content exposure detected (high risk: 0.85). Yes: the required workflow runs scripts/project_scraper.py / scripts/inventory_scan.py, which read arbitrary text from the target project files (outsider-authored code/docs) and emit it as Markdown/JSON that the agent then ingests into the LLM context as a “review pack”/inventory leads.