Skip to main content
Glama

GPT-MCP Bridge

by bzzimmy
npmGitHub
TypeScript
1
OverviewInspectNewEndpointsSchemaRelated ServersReviewsScore
Need Help?View Source CodeReport Issue

GPT-MCP Bridge

A Model Context Protocol (MCP) server that enables Claude Code and other MCP-compatible tools to communicate with OpenAI's GPT models, featuring conversation history, session management, and advanced controls.

Features

🎯 Core Capabilities

  • Conversation History - Maintain context across multiple interaction's

  • Session Management - Create, manage, and track parallel conversation sessions

  • Multi-Model Support

  • GPT-5, GPT-5-mini, and o3 models

  • Advanced Controls - Reasoning effort and verbosity parameters

  • Token Tracking - Monitor usage per session for cost management

  • Error Handling - Robust error recovery and session validation

💡 Key Benefits

  • 70% Token Savings - Reuse context without repeating information

  • Parallel Workflows - Handle mu ltiple independent tasks simultaneously

  • Adaptive Responses - Control response length with verbosity settings

  • Smart Model Routing - Choose optimal model for each task

Installation

  1. Clone the repository

git clone https://github.com/yourusername/gpt-mcp.git cd gpt-mcp

  1. Install dependencies

npm install

  1. Configure environment

cp .env.example .env # Edit .env and add your OpenAI API key

  1. Build the project

npm run build

Usage

Running the Server

Development mode:

npm run dev

Production mode:

npm start

MCP Tools

1. askGPT

Send prompts to GPT models with optional conversation context.

Parameters:

  • model (required): "gpt-5" | "gpt-5-mini" | "o3"

  • prompt (required): Your question or request

  • reasoning_effort (optional): "minimal" | "low" | "medium" | "high"

  • verbosity (optional): "low" | "medium" | "high"

  • session_id (optional): Session ID for conversation context

Example:

{ "model": "gpt-5", "prompt": "Explain async/await", "reasoning_effort": "medium", "verbosity": "low", "session_id": "abc-123" }

2. createSession

Create a new conversation session for maintaining context.

Parameters:

  • system_prompt (optional): System message to set context

Returns: Session ID for use in subsequent calls

3. clearSession

Clear a conversation session and its history.

Parameters:

  • session_id (required): Session to clear

4. listSessions

List all active conversation sessions.

Returns: Array of session information including ID, creation time, message count, and token usage

5. getSessionInfo

Get detailed information about a specific session.

Parameters:

  • session_id (required): Session to query

Configuration

Environment Variables

  • OPENAI_API_KEY - Your OpenAI API key (required)

Session Limits

  • Max tokens per session: 100,000

  • Max messages per session: 100

  • Session expiry: 24 hours

  • Auto-cleanup of expired sessions

Example Workflows

Multi-Step Debugging

// Create a session for debugging session_id = createSession("Help debug React component") // Step 1: Present the problem askGPT({ model: "gpt-5", prompt: "My component re-renders on every keystroke", session_id: session_id }) // Step 2: Ask follow-up (context maintained) askGPT({ model: "gpt-5", prompt: "How do I fix this?", session_id: session_id })

Parallel Tasks

// Create separate sessions for different tasks bugSession = createSession("Fixing memory leak") featureSession = createSession("Adding authentication") // Work on both independently askGPT({ model: "gpt-5", prompt: "...", session_id: bugSession }) askGPT({ model: "gpt-5-mini", prompt: "...", session_id: featureSession })

Model Selection Guide

GPT-5

  • Best for: Complex tasks, code generation, detailed analysis

  • Reasoning levels: minimal, low, medium, high

  • Use when: Quality is priority over speed

GPT-5-mini

  • Best for: Simple queries, quick responses, cost optimization

  • Reasoning levels: minimal, low, medium, high

  • Use when: Speed and cost are priorities

o3

  • Best for: Logic puzzles, mathematical reasoning, complex problem-solving

  • Reasoning levels: Not applicable (always maximum)

  • Use when: Deep reasoning is required

Architecture

gpt-mcp/ ├── src/ │ ├── index.ts # MCP server setup │ ├── services/ │ │ ├── openai.ts # OpenAI API integration │ │ └── session-manager.ts # Session management │ ├── tools/ │ │ ├── askGPT.ts # Main GPT interface │ │ └── session-tools.ts # Session management tools │ └── types/ │ └── index.ts # TypeScript interfaces

Development

Building

npm run build

Type Checking

npx tsc --noEmit

Project Structure

  • ES Modules - Uses .js extensions in imports

  • Strict TypeScript - Full type safety

  • MCP SDK - Built on official MCP TypeScript SDK

Requirements

  • Node.js >= 18

  • OpenAI API key

  • MCP-compatible client (Claude Code, etc.)

License

MIT

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

Support

For issues and questions, please use the GitHub issue tracker.

This server cannot be installed

-
security - not tested
F
license - not found
-
quality - not tested

How are these scores calculated?

remote-capable server

The server can be hosted and run remotely because it primarily relies on remote services or has no dependency on the local environment.

Enables Claude and other MCP-compatible tools to communicate with OpenAI's GPT models (GPT-5, GPT-5-mini, o3) with conversation history and session management. Features advanced controls like reasoning effort settings, token tracking, and parallel conversation sessions for efficient AI workflows.

  1. Features
    1. 🎯 Core Capabilities
    2. 💡 Key Benefits
  2. Installation
    1. Usage
      1. Running the Server
      2. MCP Tools
    2. Configuration
      1. Environment Variables
      2. Session Limits
    3. Example Workflows
      1. Multi-Step Debugging
      2. Parallel Tasks
    4. Model Selection Guide
      1. GPT-5
      2. GPT-5-mini
      3. o3
    5. Architecture
      1. Development
        1. Building
        2. Type Checking
        3. Project Structure
      2. Requirements
        1. License
          1. Contributing
            1. Support

              New MCP Servers

              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/bzzimmy/gpt-mcp'

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