llm-prompt-optimizer

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

• Third-party content exposure detected (high risk: 0.75). SKILL.md contains the skill’s imported editorial workflow text (including “Imported: …” sections) that becomes LLM-readable context at runtime, and because it is packaged from an external public repository (ORIGIN.md shows GitHub community source), that content is outsider-authored free text used to form prompts.