edge-hint-extractor

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

• Third-party content exposure detected (high risk: 0.80). The skill's build_hints.py (and SKILL.md / hints_schema.md) documents and implements an --llm-ideas-cmd that runs an arbitrary external LLM command and ingests/parses its stdout as hints which are merged into the generated hints.yaml, meaning untrusted third‑party LLM output can directly influence the tool's decisions.