The Agent Skills Directory

EXTERNAL_DOWNLOADS (SAFE): Dependencies are restricted to reputable packages (google-genai, Pillow) from trusted maintainers. No suspicious external script downloads were detected.- CREDENTIALS_UNSAFE (SAFE): The skill correctly manages authentication using the GEMINI_API_KEY environment variable. No hardcoded secrets or embedded API keys were identified.- COMMAND_EXECUTION (SAFE): No evidence of arbitrary command execution or unsafe shell usage. All functionality is implemented through standard library calls and official SDK methods.- REMOTE_CODE_EXECUTION (SAFE): No patterns of remote script execution, dynamic code loading, or unsafe deserialization were found.

gemini-imagegen