dual-axis-skill-reviewer

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

• Third-party content exposure detected (high risk: 0.85). The skill reads and ingests OUTSIDER-authored free text from the target project’s skills/*/SKILL.md (and related scripts/*.py / references/*.md) via score_skill() → skill_file.read_text() and child.read_text(), then embeds that content into the generated LLM prompt in build_llm_prompt() (indirect prompt injection risk).