skill-composition

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

• Third-party content exposure detected (medium risk: 0.30). The required runtime workflow ingests outsider-authored free text via the query field and passes it to LLM semantic matching/dependency analysis (e.g., scripts/skill_discovery.py builds user_prompt with query, then calls llm_chat, and scripts/workflow_composer.py includes query in the dependency-analysis prompt and calls llm_chat).