Awesome MCP › Software Development

bazinga012/mcp_code_executor

⭐ 211 JavaScript added to this list on 2025-05-14 repository created 2025-02-06

MCP Code Executor is a specialized MCP (Model Context Protocol) server designed to enable large language models (LLMs) to execute Python code within controlled and specified Python environments. This project supports incremental code generation, which is particularly useful for handling large code blocks that might exceed token limits in LLM prompts. The server allows LLMs to run Python code with access to libraries and dependencies defined in environments such as Conda, standard Python virtualenv, or UV virtualenv, providing flexibility in environment management. Key features include the ability to execute Python code snippets, install and check for Python package dependencies dynamically, and configure the execution environment at runtime. The server supports a range of tools that facilitate code execution workflows, including initializing new Python code files, appending code to existing files, reading file contents, and executing complete Python scripts. This modular approach enables LLMs to build and run complex multi-part Python programs incrementally. The MCP Code Executor is designed to be integrated into MCP server configurations, with support for running via Node.js or Docker. It requires setting environment variables to specify the code storage directory and environment type, along with relevant environment details such as Conda environment name or virtualenv paths. The project emphasizes backward compatibility, ensuring that users with existing Conda environment setups can upgrade without configuration changes. Overall, MCP Code Executor enhances the capabilities of LLMs by providing a robust and flexible infrastructure for executing Python code within defined environments, supporting dynamic dependency management and incremental code generation to overcome token limitations. This makes it a valuable tool for applications requiring programmatic code execution and environment control within the MCP ecosystem.

https://github.com/bazinga012/mcp_code_executor

append_to_code_filecheck_installed_packagescode-executioncode-file-managementcode-storagecondaconda-environmentconfigure_environmentdependency-installationdependency-managementdockerdynamic-environmentdynamic-environment-managementenvironment-configurationexecute_codeexecute_code_fileget_environment_configincremental-code-generationinitialize_code_fileinstall_dependencieslarge-language-modelsllmmcpmcp-servermodel-context-protocolnode.jspython-code-executionpython-environmentpython-packagesread_code_filetoken-limit-handlingtoken-limitsuv-virtualenvvirtualenv