seo-image-gen

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

• Third-party content exposure detected (high risk: 0.80). The skill's SKILL.md explicitly says an audit agent is spawned during "/seo audit" to "analyze OG/social images across the site" (reading site pages) and the references (references/gemini-models.md and references/prompt-engineering.md) document and exemplify Google Search / search-grounded generation, indicating the agent ingests public web/image search results and site pages which could be untrusted and then uses those findings to drive generation decisions.