skill-cleaner

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

• Third-party content exposure detected (high risk: 0.75). The required runtime workflow reads outsider-authored free text from local skill files and recent log files (e.g., SKILL.md bodies and session/history .jsonl/.log text) via fs.readFileSync(..., "utf8"), then renders that content into the report output that is subsequently consumed by the agent/LLM context.