AI coding assistants are powerful but have limitations—training data cuts off at a specific date, missing new API features and changes. Without access to Gemini-specific documentation, agents may suggest generic patterns instead of optimized approaches.
To keep your coding assistant current with the evolving Gemini API and its recommended usage, we recommend setting up the Gemini Docs MCP and enhancing your environment with Gemini API Skills. While these tools are usable independently, they are designed to work together to provide complete coverage.
Connect the Gemini Docs MCP
Gemini hosts a public Model Context Protocol (MCP) server at
gemini-api-docs-mcp.dev. Connecting your coding agent to this server ensures that
all queries have access to the latest APIs, code updates, and optimal configuration examples.
Run the following command in your agent's terminal or project root to install the server:
npx add-mcp gemini-api-docs-mcp.dev
This server adds a search_documentation function that your agent can use to
retrieve real-time API definitions and integration patterns from the official
Gemini documentation files.
Add API Development Skills
The skills provide baked-in rules and best practices (such as enforcing the correct SDK and current model versions) directly in your assistant's context. The skill works together with the Gemini Docs MCP service: If you have both
installed, the skill uses the MCP service for documentation, but even without
the MCP installed, it will fetch llms.txt from ai.google.dev as a fallback.
To install these skills, you can use one of the following supported tools. Installation instructions for both are provided below each skill module:
- skills.sh: Recommended. The open standard for portable agent behaviors.
- Context7: Supported for users already utilizing the Context7 ecosystem.
gemini-api-dev
The foundational skill for general-purpose Gemini development. This skill provides documentation and best practices for:
- Prompt routing to current models (e.g., Gemini 3.1 Pro/Flash) and avoiding deprecated models
- Multimodal prompting, function calling, structured outputs, and common integration patterns
Install with skills.sh
npx skills add google-gemini/gemini-skills --skill gemini-api-dev --global
Install with Context7
npx ctx7 skills install /google-gemini/gemini-skills gemini-api-dev
gemini-live-api-dev
Skill for building real-time conversational AI applications with Gemini Live API. This skill provides documentation and best practices for:
- WebSocket connections for low-latency streaming
- Streaming audio, video, and text
- Voice activity detection and barge-in support
Install with skills.sh
npx skills add google-gemini/gemini-skills --skill gemini-live-api-dev --global
Install with Context7
npx ctx7 skills install /google-gemini/gemini-skills gemini-live-api-dev
gemini-interactions-api
Skill for building apps with the Interactions API. The Interactions API is a unified interface for interacting with Gemini models and agents, designed for agentic applications. This skill covers:
- Text generation, multi-turn chat, and streaming
- Function calling, structured output, and image generation
- Background execution and Deep Research agents
- Server-side conversation state management
- Python and TypeScript SDK patterns
Install with skills.sh
npx skills add google-gemini/gemini-skills --skill gemini-interactions-api --global
Install with Context7
npx ctx7 skills install /google-gemini/gemini-skills gemini-interactions-api
Verify installation
After installing, confirm that your coding assistant can connect to the Gemini Docs MCP server and use your installed skills.
1. Verify agent behavior
The most reliable way to verify is to ask your agent a technical question about Gemini API.
Prompt: "How do I use context caching with the Gemini API?"
A successful setup will:
- Provide accurate code: Reference specific Gemini methods like
cacheContentorcachedContents.createfrom the latest endpoints. - Use the MCP Tool: Show that it is connected to the Gemini Docs MCP Server or utilizing the
search_documentationtool to fetch data. - Invoke loaded skills: Show an indicator that it is "Using skill: gemini-api-dev" (if relying on a secondary wrapper).
2. Verify manifestations & tools
If the agent gives a general or generic answer, use the specific Discovery or Status commands for your environment to verify that the Docs MCP or skill is loaded into memory.
| Environment | MCP Verification | Skills Verification |
|---|---|---|
| Claude Code | Type /mcp in the terminal to view active servers and search_documentation tools. |
Type /skills in the terminal to list all active manifests. |
| Cursor | Navigate to Settings > Features > MCP. Ensure server is "Connected". | Open Settings > Rules. Verify the skill appears under "Agent Decides." |
| Antigravity | Check the Customizations > Connections sidebar for MCP status. | Type /skills list or check the Customizations > Rules sidebar. |
| Gemini CLI | Run gemini mcp list or use /mcp list. |
Run gemini skills list or use the /skills slash command in-session. |
| Copilot | Type @gemini /mcp to list active data connectors. |
Type @gemini /skills (or /skills) to view active extensions. |
Troubleshooting
If your agent provides only general information or fails to recognize Gemini-specific methods, check the following:
Agent didn't discover the skill
Most agents index skills only on startup.
Fix: Completely restart your IDE (Cursor/VS Code) or exit and re-open your terminal-based agent (Claude Code).
Global vs. local conflict
If you installed with the --global flag, your agent might be ignoring it in
favor of project-specific rules.
Fix: Try installing the skill directly into your project root without the global flag:
npx skills add google-gemini/gemini-skills --skill gemini-api-dev