Azure Code Interpreter MCP

What it does\nConnects AI agents to Azure Code Interpreter, allowing them to execute Python code in a secure, isolated, and ephemeral sandboxed environment. This enables complex data analysis and file generation without risking the local host system.\n\n## Tools\n- new_session: Creates a new isolated execution session to delineate different tasks.\n- execute: Runs Python code within the current session. Files are saved to /mnt/data for persistence and download.\n- list_files: Lists all files stored in the /mnt/data directory produced by executed code.\n- download_files: Transfers files from the Azure sandbox to a local directory specified by AZURE_DOWNLOAD_DIRECTORY.\n\n## Installation\nBuild the binary using Go and add it to your PATH. Then add to your config:\n\njson\n{\n "mcpServers": {\n "azure-code-interpreter": {\n "command": "azure-code-interpreter-mcp",\n "env": {\n "AZURE_SUBSCRIPTION_ID": "your-sub-id",\n "AZURE_RESOURCE_GROUP": "your-rg",\n "AZURE_SESSION_POOL": "your-pool",\n "AZURE_REGION": "your-region",\n "AZURE_DOWNLOAD_DIRECTORY": "/path/to/downloads"\n }\n }\n }\n}\n\n\n## Supported hosts\n- Claude Desktop