Context Bank MCP

Context Bank MCP

An MCP server that provides an interface for querying the AtherOS knowledge base through an API, allowing users to create chat sessions and send queries to retrieve information.

Category
Visit Server

README

Context Bank MCP

Overview

Context Bank MCP is a project using the Model Context Protocol (MCP) to create an interface for querying the AtherOS knowledge base through an API. This project builds an MCP server capable of interacting with the Onyx API to create chat sessions and send queries to the knowledge base.

Features

  • Create new chat sessions to query the knowledge base
  • Send messages to chat sessions to receive responses from the knowledge base
  • Format and display results from the Onyx API

Technologies Used

  • TypeScript
  • Node.js
  • Model Context Protocol (MCP) SDK
  • Zod for data validation
  • Axios for HTTP requests

Installation

# Install dependencies
npm install

# Compile source code
npm run build

Configuration

The project uses environment variables to connect to the Onyx API. Follow these steps to configure your environment:

Automatic Setup (Recommended)

Run the setup script to configure your environment interactively:

npm run setup

This script will:

  1. Create a .env file if it doesn't exist
  2. Prompt you for your AtherOS API key
  3. Allow you to customize the API base URL
  4. Set default values for other configuration options

Manual Setup

  1. Copy the example environment file to create your own:
cp .env.example .env
  1. Edit the .env file with your specific configuration:
# AtherOS API Configuration
ONYX_API_KEY=your_api_key_here
ONYX_API_BASE=http://your_api_base_url:port

# Server Configuration
PORT=3000
NODE_ENV=development

# Optional: Logging Configuration
LOG_LEVEL=info

Environment Variables

  1. Required Environment Variables:

    • ONYX_API_KEY: Your AtherOS API key
    • ONYX_API_BASE: Base URL for the AtherOS API (e.g., "http://172.30.22.52:3000")
  2. Optional Environment Variables:

    • PORT: Port number for the server (default: 3000)
    • NODE_ENV: Environment mode (development, production, test)
    • LOG_LEVEL: Logging level (info, debug, error, warn)

Usage

After compilation, you can use the command line tool:

# Direct usage
./build/index.js

# Or through npm
npm start

API Tools

The project provides two main MCP tools:

1. create_chat_session

Creates a new chat session to query the knowledge base.

Parameters:

  • persona_id (default: 0): User ID
  • description (default: ""): Chat session description

2. query_atheros

Sends a message to the chat session to query the AtherOS knowledge base.

Parameters:

  • chat_session_id: Chat session ID
  • message: Message content
  • parent_message_id: Parent message ID (can be null)

Response Format

Responses from knowledge base queries include:

  • Message ID
  • Message content
  • Rephrased query (if available)
  • Information about top source documents (if available):
    • Document name
    • Relevance score
    • Link to the document

Architecture

The project is organized with a simple structure:

  • src/index.ts: Main entry point of the application, defines the MCP tools and connection logic
  • Uses the stdio protocol to communicate with the MCP server

Development

# Compile and view changes
npm run build

# Run in development mode
npm run dev

License

ISC

Recommended Servers

playwright-mcp

playwright-mcp

A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.

Official
Featured
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.

Official
Featured
Local
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.

Official
Featured
Local
TypeScript
VeyraX MCP

VeyraX MCP

Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.

Official
Featured
Local
graphlit-mcp-server

graphlit-mcp-server

The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.

Official
Featured
TypeScript
Kagi MCP Server

Kagi MCP Server

An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.

Official
Featured
Python
E2B

E2B

Using MCP to run code via e2b.

Official
Featured
Neon Database

Neon Database

MCP server for interacting with Neon Management API and databases

Official
Featured
Exa Search

Exa Search

A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.

Official
Featured
Qdrant Server

Qdrant Server

This repository is an example of how to create a MCP server for Qdrant, a vector search engine.

Official
Featured