gemini-api-dev

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 1.00). The skill explicitly instructs at runtime to fetch the index and documentation pages (e.g., https://ai.google.dev/gemini-api/docs/llms.txt and linked .md.txt pages) as a required fallback when MCP is unavailable, and that fetched content is used to determine API behavior/instructions, so it directly controls the agent's prompts.