Connects to Google Gemini's image generation and editing APIs through the Model Context Protocol. You get programmatic access to Gemini's visual AI capabilities without building your own REST client or handling auth flows. Useful when you want to generate or modify images from within an LLM workflow, whether that's creating variations, applying edits, or producing new visuals from text prompts. Part of Mindstone's 35-connector suite, so it follows the same stdio transport pattern and configuration style as their Google Workspace, Slack, and Microsoft integrations. Ships as an npm package under the @mindstone scope with build-time security controls and weekly dependency cool-downs.
claude mcp add --transport stdio io.github.mindstone-mcp-server-nano-banana uvx mcp-server-nano-banana