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 explicitly accepts and runs external LLM output (via --llm-ideas-cmd subprocess and by loading --llm-ideas-file YAML) as part of its required workflow (see SKILL.md quick commands and scripts/build_hints.py generate_llm_hints / load_llm_hints_from_file), so untrusted, third-party-generated content is ingested and directly influences the produced hints used downstream.