Skip to main content
Glama

Gemini MCP Tool

sandbox.md1.96 kB
# Sandbox Mode Execute code safely in an isolated environment. ## What is Sandbox Mode? Sandbox mode allows Gemini to write and test code in a secure, isolated environment without affecting your system. ## Basic Usage ``` /gemini-cli:sandbox create a Python script that sorts a list ``` ## How It Works 1. **Request** → You ask for code to be created/tested 2. **Generation** → Gemini writes the code 3. **Execution** → Code runs in isolated environment 4. **Results** → Output returned safely ## Use Cases ### Algorithm Testing ``` /gemini-cli:sandbox implement and test quicksort in JavaScript ``` ### Data Processing ``` /gemini-cli:sandbox parse this CSV and show statistics: [data] ``` ### Proof of Concepts ``` /gemini-cli:sandbox create a working web scraper example ``` ## Safety Features - **Isolated Execution**: No access to your file system - **Resource Limits**: CPU and memory constraints - **Time Limits**: Prevents infinite loops - **No Network**: Cannot make external requests ## Supported Languages - Python - JavaScript/Node.js - Ruby - Go - Java - C++ - More coming soon! ## Best Practices ### 1. Be Specific ``` // Good create a function that validates email addresses with tests // Vague make something that checks emails ``` ### 2. Include Test Cases ``` implement binary search with edge case handling and show test results ``` ### 3. Iterative Development ``` // First iteration create a basic REST API // Refine add authentication to the API // Test show example requests and responses ``` ## Limitations - No file system access - No network requests - Limited execution time (30s) - Memory limit (512MB) ## Examples ### Testing Algorithms ``` /gemini-cli:sandbox benchmark bubble sort vs quick sort with 1000 items ``` ### Learning Code ``` /gemini-cli:sandbox show me how promises work in JavaScript with examples ``` ### Debugging ``` /gemini-cli:sandbox why does this code fail: [paste code] ```

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/jamubc/gemini-mcp-tool'

If you have feedback or need assistance with the MCP directory API, please join our Discord server