visualize

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

• Third-party content exposure detected (medium risk: 0.65). SKILL.md describes reading evaluation data from ~/.claude/.skill-evaluator/skills/{YYYY-MM}.json and scanning installed skills’ metadata, which can include outsider-authored free text (e.g., other skills’ description/trigger fields) that is then injected into the HTML template via __SKILL_UP_DATA__ and thus becomes LLM-readable context at runtime.