seo-images

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

• Third-party content exposure detected (high risk: 0.75). SKILL.md describes a runtime workflow that fetches Google Images SERP results (“fetch Google Images results via serp_google_images_live_advanced”, then extracts top domains/alt patterns), which would ingest free-form text from third-party web content into the agent’s LLM context for analysis; the category is “Public web content fetched at runtime — pages/articles/search results”.

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 1.00). The skill explicitly recommends running privileged package-install commands (e.g., "sudo apt install libimage-exiftool-perl" and "sudo apt install webp"), which pushes the agent toward obtaining sudo privileges and modifying the host system state.