plugin-maintain

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

• Third-party content exposure detected (high risk: 0.72). Workflow “add-knowledge” (SKILL.md:145-215) can fetch outsider-authored free text from a user-supplied URL at runtime (“If URL: Use WebFetch to retrieve content”), then converts it to Markdown and writes it into the skill’s reference files—making that outsider text available for later LLM context loading.