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 tells the agent to use fetch_url at runtime to retrieve https://ai.google.dev/gemini-api/docs/llms.txt (and the linked https://ai.google.dev/gemini-api/docs/*.md.txt pages) as the authoritative documentation to drive its behavior, meaning externally fetched content would directly control prompts/instructions.