gemma-dev

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly directs the agent to use fetch_url at runtime to retrieve docs (e.g., https://ai.google.dev/gemma/docs/llms.txt and https://ai.google.dev/gemma/docs/core/prompt-formatting-gemma4.md.txt), and those fetched pages are used to control prompt formatting and API behavior, so this is a runtime external dependency that can directly control agent prompts.