MCP Server Hub

A Python-based Model Context Protocol (MCP) server that exposes tools for multiple integrations — GitHub, Jira, REST API, Oracle, and MongoDB — for use by AI assistants.

Prerequisites

• Python 3.10+
• Credentials for the integrations you want to use (see Configuration)

Installation

pip install -e .

Browser UI

Start the server in SSE mode and open the browser UI to explore and test all tools interactively:

python -m mcp_server --sse

Then open: http://localhost:8000/ui

Tools are grouped by integration in the sidebar. Select a tool, fill in the parameters, and click Run.

Configuration

Each integration is activated by setting its environment variables. Integrations without credentials are gracefully skipped at startup.

GitHub

export GITHUB_TOKEN="ghp_your_token_here"

Jira

export JIRA_URL="https://your-org.atlassian.net"
export JIRA_USERNAME="you@example.com"
export JIRA_API_TOKEN="your_api_token"

REST API

export REST_API_BASE_URL="https://api.example.com"
export REST_API_TOKEN="your_token"           # optional
export REST_API_TOKEN_HEADER="Authorization" # optional, default: Authorization
export REST_API_TOKEN_PREFIX="Bearer"        # optional, default: Bearer

Oracle Database

export ORACLE_USER="your_user"
export ORACLE_PASSWORD="your_password"
export ORACLE_DSN="host:port/service_name"

MongoDB

export MONGODB_URI="mongodb://localhost:27017"
export MONGODB_DATABASE="your_database"

On Windows (PowerShell), use $env:VAR = "value" syntax.

Usage

Run directly (stdio — for MCP clients)

mcp-server-hub

Run with browser UI (SSE mode)

mcp-server-hub --sse
# or
python -m mcp_server --sse

VS Code / Copilot

Add to your .vscode/mcp.json:

{
  "servers": {
    "hub": {
      "type": "stdio",
      "command": "python",
      "args": ["-m", "mcp_server"],
      "env": {
        "GITHUB_TOKEN": "ghp_your_token_here",
        "JIRA_URL": "https://your-org.atlassian.net",
        "JIRA_USERNAME": "you@example.com",
        "JIRA_API_TOKEN": "your_api_token",
        "MONGODB_URI": "mongodb://localhost:27017",
        "MONGODB_DATABASE": "your_database"
      }
    }
  }
}

Claude Desktop

Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "hub": {
      "command": "python",
      "args": ["-m", "mcp_server"],
      "env": {
        "GITHUB_TOKEN": "ghp_your_token_here"
      }
    }
  }
}

Available Tools

GitHub (11 tools)

Tool	Description
github_search_repositories	Search for GitHub repositories
github_get_repository	Get detailed info about a repository
github_list_issues	List issues in a repository
github_get_issue	Get details of a specific issue
github_create_issue	Create a new issue
github_list_pull_requests	List pull requests in a repository
github_get_pull_request	Get details of a specific pull request
github_get_file_contents	Get file or directory contents from a repo
github_list_branches	List branches in a repository
github_list_commits	List commits in a repository
github_search_code	Search for code across GitHub

Jira (8 tools)

Tool	Description
jira_list_projects	List all accessible Jira projects
jira_search_issues	Search issues using JQL
jira_get_issue	Get full details of an issue
jira_create_issue	Create a new issue
jira_update_issue	Update fields of an existing issue
jira_add_comment	Add a comment to an issue
jira_list_transitions	List available workflow transitions
jira_transition_issue	Transition an issue to a new status

REST API (5 tools)

Tool	Description
rest_get	Make an HTTP GET request
rest_post	Make an HTTP POST request
rest_put	Make an HTTP PUT request
rest_patch	Make an HTTP PATCH request
rest_delete	Make an HTTP DELETE request

Oracle Database (5 tools)

Tool	Description
oracle_execute_query	Execute a SELECT query and return results as JSON
oracle_execute_statement	Execute an INSERT, UPDATE, or DELETE statement
oracle_list_tables	List tables in the database
oracle_describe_table	Describe columns of a table
oracle_call_procedure	Call a stored procedure

MongoDB (9 tools)

Tool	Description
mongo_list_collections	List all collections in the database
mongo_find	Find documents with a filter
mongo_find_one	Find a single document
mongo_insert_document	Insert one document
mongo_insert_many	Insert multiple documents
mongo_update_document	Update a document
mongo_delete_document	Delete a document
mongo_count	Count documents matching a filter
mongo_aggregate	Run an aggregation pipeline

Adding a New Integration

1. Create mcp_server/tools/<name>/ with client.py and tools.py
2. Decorate each tool function with @tool("Category Name") from mcp_server.registry
3. Add the module import to mcp_server/tools/__init__.py

The tool will automatically appear in the browser UI and be available to MCP clients.