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 docs explicitly instruct and support Google Search grounding and web/image search via the MCP tools (see references/gemini-models.md "Google Search Grounding", references/prompt-engineering.md "Search-Grounded Generation", and references/mcp-tools.md feature list), meaning the agent can fetch and analyze open/web image and page content and incorporate that untrusted third‑party material into its prompt/decision workflow.