Complete your Gueini Cli preparation guide for your station

Gemini Cli has some compact tools within reach. You can use orders such as:

Overview of Gemini Cli tools

• reading ((ls). It lists files and folders in a guide – just like the shell ls He orders.
• Readfile ((read-file). Read the full content of one file, useful for summaries or analysis.
• Readanyfiles ((read-many-files). He reads many files simultaneously, usually matchs the globe style (for example, everything .js Files)
• Findfiles ((glob). Looking for files by style (for example, looking for everyone config.json Through your project).
• Searchtext ((grep). Looking inside the files of the text, such as finding all TODO comments.
• Restricting ((edit). Code changes are applied via Diffs. Gemini and ask about approval before applying it.
• Writing ((write-file). It creates new files (for example, README.md) With the content offered by the user.
• coincidence ((shell). He manages the orders he earns with ! (For example , !npm test) Directly at the station.
• Webfetch ((web-fetch). It attends the web content (HTML or JSON), enabling Gemini to analyze external data.
• Googlesearch ((web-search). Google research leads to ground responses with information in the real world (for example, an explanation of a mistake).
• Memory save ((memoryTool). Store facts or preferences during the session (such as “I am the best synchronous/waiting”) to improve consistency

To see all available tools, you can use /tools He orders.

Advanced features

You can add specific instructions to the IQ Agency for a specific project by creating a file Gemini.md A file in the root guide for your project. Inside this file, you can determine the project rules, code patterns and tools that the agent must use. This ensures that the created code corresponds to your project standards.

Google Cli MCP integration

For most daily uses, compact tools are sufficient. But what if you want to do Gemini Cli something very special in the field, such as interacting with specific application programming facades or using a specialized model (less photo generator or safety analysis tool)? This is the place MCP (Form Context Protocol) It comes in.

MCP is mainly an open standard that allows developers Add new tools/capabilities to artificial intelligence by running a server The Cli can communicate with. In Gemini Cli, you can configure “MCP servers” in the JSON settings file, and Cli will treat those additional tools you can use.

How to prepare a MCP server in Google Cli

For example, I will show you how to prepare GitHub McP Server in Gemini Cli.

Inside your project folder, create a folder using:

mkdir -p .gemini && touch .gemini/settings.json

Inside the file, add the following symbol:

{  
  "mcpServers": {  
    "github": {  
      "command": "npx",  
      "args": ["-y", "@modelcontextprotocol/server-github"],  
      "env": { "GITHUB_PERSONAL_ACCESS_TOKEN": "[YOUR-TOKEN]" }  
    }  
  }  
}

after that /quit From Gemini Cli, then reopen it.

He writes /mcp The matter, and you will see a list of Gaytap tools.

Now you, an agent, you can interact with Gyfap. This is simple! Prevailing

Gemini Cli pricing

You can try it for free for personal use, but there is also a paid version based on the use of the distinctive symbol.

Free preview layer

• Assign: Free with Google account
• border: until 60 requests/min and 1000 requests/day
• Wonderful for individual developers and use on a small scale

Use the paid API key / API

• Use yourself Gemini key API For higher use
• Bills based on consuming codes (based on form and use)

Foundation options

• Available Help the Gemini icon or the institution Plan
• It includes advanced features such as governance, audit records and shared shares

conclusion

As you can see, Gemini Cli is a truly powerful tool with a lot of capabilities. I am excited to see how I will use it in the daily workflow.

If you write a code, correct things or manage files often, this tool is worth verifying.

If you have any notes, please share it in the comments below. 😉

Cheers!