MCP Coding Assistant with support for OpenAI + other LLM Providers

Name: MCP Coding Assistant with support for OpenAI + other LLM Providers
Author: arthurcolle

@arthurcolle

リポジトリ

MCP Coding Assistant with support for OpenAI + other LLM Providers について

OpenAI Code Assistant Model Context Protocol (MCP) Server

基本情報

カテゴリ

AI とエージェント

ランタイム

python

トランスポート

stdio

公開者

arthurcolle

設定

以下の設定を使って、このサーバーを MCP 対応クライアントに追加してください。

{
  "mcpServers": {
    "openai-mcp": {
      "command": "python",
      "args": [
        "claude.py",
        "chat"
      ]
    }
  }
}

ツール

** Read files with optional line limits

** Modify files with precise text replacement

** Create or overwrite files

** Find files by pattern matching

** Search file contents using regex

** List directory contents

** Execute shell commands

概要

What is MCP Coding Assistant with support for OpenAI + other LLM Providers?

A Python-based coding assistant that provides a natural language interface for software development tasks, supporting multiple LLM providers (OpenAI, Anthropic). It includes Model Context Protocol (MCP) server and client capabilities, real-time tool visualization, cost management, and multi-agent coordination.

How to use MCP Coding Assistant with support for OpenAI + other LLM Providers?

Clone the repository, install dependencies with pip install -r requirements.txt, and set API keys in a .env file. Run in CLI mode with python claude.py chat, as an MCP server with python claude.py serve, as an MCP client with python claude.py mcp-client path/to/server.py, or in multi-agent mode with python claude.py mcp-multi-agent path/to/server.py --config config.json. Use flags like --provider, --model, and --budget to configure behavior.

Key features of MCP Coding Assistant with support for OpenAI + other LLM Providers

Multi-provider support (OpenAI, Anthropic)
Run as MCP server for Claude Desktop clients
Built-in MCP client to connect to any MCP server
Multi-agent synchronization for complex problem solving
Real-time tool execution progress and results
Cost tracking and budget controls
Tools: View, Edit, Replace, GlobTool, GrepTool, Bash, LS
Chat commands for compact, cost, budget, and providers

Use cases of MCP Coding Assistant with support for OpenAI + other LLM Providers

Interactive coding assistance via CLI with multiple LLM providers
Using the assistant as an MCP server for integration with Claude Desktop
Connecting to external MCP servers for tool augmentation
Collaborative problem-solving with specialized agents in multi-agent mode
Controlled-cost development sessions with budget limits

FAQ from MCP Coding Assistant with support for OpenAI + other LLM Providers

Which LLM providers are supported?

OpenAI and Anthropic are supported; configure via OPENAI_API_KEY and ANTHROPIC_API_KEY environment variables.

What are the system requirements?

Python 3.10 or higher, with dependencies installed from requirements.txt. API keys for at least one provider must be set.

How do I manage costs?

Use the --budget flag when starting a session (e.g., --budget 5.00), and use the /cost and /budget chat commands to track and adjust spending.

What tools are available in the assistant?

View, Edit, Replace, GlobTool, GrepTool, LS, and Bash. These allow file reading, modification, search, pattern matching, directory listing, and command execution.

Can I use this as a server for other MCP clients?

Yes, run python claude.py serve to start the MCP server. The server provides a configuration page at http://localhost:8000 with instructions for connecting Claude Desktop.