IBM i MCP Server
Enables AI agents to interact with IBM i systems through the Model Context Protocol, providing secure SQL-based access to Db2 for i databases for querying, monitoring, and executing operations.
README
<div align="center">

MCP server and CLI for IBM i </div>
Overview
The IBM i MCP Server enables AI agents to interact with IBM i systems through the Model Context Protocol (MCP). It provides secure, SQL-based access to Db2 for i databases, allowing AI applications like Claude, VSCode Copilot, Bob, and custom agents to query system information, monitor performance, and execute database operations.
This repo also ships the ibmi CLI (@ibm/ibmi-cli) β a terminal-first sibling that shares the same YAML-driven SQL tool engine, so the same tool definitions work in both the MCP server and the CLI.

How it works: AI clients connect via MCP β Server executes YAML-defined SQL tools β Results stream back to the AI agent through Mapepire.
[!TIP] π Official Documentation | β οΈ Docs are under active development
The Docs are continuously evolving. Please check back frequently for updates and new guides. If there's something missing, feel free to open an issue!
Repository Structure
| Directory | Purpose | Documentation |
|---|---|---|
packages/server/ |
MCP server implementation (TypeScript) β @ibm/ibmi-mcp-server on npm |
Server README |
packages/cli/ |
ibmi command-line interface β @ibm/ibmi-cli on npm, co-versioned with the server |
CLI README |
tools/ |
YAML-based SQL tool configurations | Tools Guide |
agents/ |
AI agent examples and integrations | Agents Guide |
client/ |
Python client examples for testing | Client README |
deployment/ |
Docker, Podman, OpenShift configs | Deployment Guide |
app/ |
Example Full stack Agent app (AgentOS) | App README |
Quick Navigation
- MCP Server - Get started with the server
- IBM i CLI - Terminal-first SQL and tool execution
- SQL Tools - Create custom SQL tools
- AI Agents - Use agent frameworks
- Python Clients - Test with Python clients
- Deployment - Deploy to production
- Setup Mapepire - Install prerequisite
MCP Server
The MCP Server enables AI agents to execute SQL queries on IBM i systems through YAML-defined SQL tools.
Quick Start
Prerequisites:
- Mapepire running on IBM i (port 8076)
- Node.js 18+ installed
Get Started:
-
Clone the repository:
git clone https://github.com/IBM/ibmi-mcp-server.git cd ibmi-mcp-server -
Configure your IBM i connection:
cat > .env << 'EOF' DB2i_HOST=your-ibmi-host.com DB2i_USER=your-username DB2i_PASS=your-password DB2i_PORT=8076 DB2i_IGNORE_UNAUTHORIZED=true EOF -
Start the server:
npx -y @ibm/ibmi-mcp-server@latest \ --transport http \ --tools ./tools/performance/performance.yamlThe server will use our pre-configured tools for:
- π Performance monitoring (system status, memory pools, active jobs)
- See the Tools Guide for more toolsets.
The MCP server can also run in a Docker container:
docker run --rm --name ibmi-mcp-server \ -v /path/to/tools/:/tools \ -v /path/to/.env/:/.env \ -e MCP_SERVER_CONFIG=/.env \ -p 3010:3010 ghcr.io/ibm/ibmi-mcp-server:latestReplace the volume paths with your actual local paths to the tools directory and
.envfile. -
Verify it's running:
# Check server health curl http://localhost:3010/healthz # List available tools curl -X POST http://localhost:3010/mcp \ -H "Content-Type: application/json" \ -H "Accept: application/json, text/event-stream" \ -d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}' | \ grep -o '"name":"[^"]*"' | sed 's/"name":"//g' | sed 's/"//g' | head -20
[!NOTE] π Full Server Quickstart β
Next Steps:
- Create Custom Tools - Build your own SQL tools
- Connect AI Clients - Integrate with Claude, VSCode, Bob, etc.
Additional links:
- Server README
- Server Configuration
- Looking for the terminal CLI instead? Jump to IBM i CLI.
IBM i CLI
A terminal-first sibling to the MCP server. The ibmi command shares the same YAML-driven SQL tool engine and is designed for local exploration, ad-hoc queries, scripted automation, and CI/CD β no MCP client required.
Quick Start
# Install
npm i -g @ibm/ibmi-cli
# Run a query
ibmi sql "SELECT * FROM SAMPLE.EMPLOYEE FETCH FIRST 5 ROWS ONLY"
# Run a YAML-defined tool
ibmi tool system_status --tools ./tools/work-management.yaml
When to use the CLI vs. the MCP Server
Both tools sit on top of the same YAML tool engine. The MCP Server is the de facto interface for building AI agents and multi-client AI workloads. The CLI covers everything else β and also serves as a lightweight, process-local alternative when a full MCP server is overkill.
| Use the CLI when⦠| Use the MCP Server when⦠|
|---|---|
| You need developer ergonomics β an interactive terminal experience for querying Db2 for i, exploring schemas, scripting in shell/CI, or piping results into other tools | You are building AI agents or AI-powered applications that should call IBM i tools conversationally (Claude Desktop, VSCode Copilot, Bob, Agno, LangChain, custom agents) |
You are running a local AI agent and want the CLI as a lightweight, in-process alternative to a long-lived MCP server β e.g., a coding agent that shells out to ibmi sql or ibmi tool as part of its loop |
You need remote server support over the MCP protocol, including stdio or HTTP transports |
You want fast synchronous execution against one or many systems with rich output formats (table, json, csv, markdown, NDJSON) and no server process to manage |
You need shared, networked access with auth, rate limiting, structured telemetry, and session handling |
See the CLI Agent Integration guide for concrete examples of wiring the
ibmiCLI into local AI agents.
[!NOTE] π Full Documentation: CLI Guide β
Additional links:
SQL Tools
YAML-based SQL tool configurations that define what queries AI agents can execute on your IBM i system.
Quick Start
Create a custom tool file tools/my-tools.yaml:
sources:
my-system:
host: ${DB2i_HOST}
user: ${DB2i_USER}
password: ${DB2i_PASS}
port: 8076
ignore-unauthorized: true
tools:
system_status:
source: ibmi-system
description: "Overall system performance statistics with CPU, memory, and I/O metrics"
parameters: []
statement: |
SELECT * FROM TABLE(QSYS2.SYSTEM_STATUS(RESET_STATISTICS=>'YES',DETAILED_INFO=>'ALL')) X
toolsets:
performance:
tools:
- system_status
Run the server with your tools:
npx -y @ibm/ibmi-mcp-server@latest --tools ./tools/my-tools.yaml --transport http
Available Tool Collections
The tools/ directory includes ready-to-use configurations:
- Performance Monitoring - System status, active jobs, CPU/memory metrics
- Security & Audit - User profiles, authorities, security events
- Job Management - Active jobs, job queues, subsystems
- Storage & IFS - Disk usage, IFS objects, save files
- Database - Tables, indexes, constraints, statistics
[!NOTE] π Full Documentation: Tools Guide β
Additional links:
AI Agents
Pre-built AI agent examples using popular frameworks to interact with IBM i systems through the MCP Server.
Available Agent Frameworks
| Framework | Language | Use Case | Documentation |
|---|---|---|---|
| Agno | Python | Production-ready agents with built-in observability | Agno README |
| LangChain | Python | Complex workflows and tool chaining | LangChain README |
| Google ADK | Python | Google AI ecosystem integration | Google ADK README |
What Agents Can Do
- System Monitoring: Real-time performance analysis and health checks
- Troubleshooting: Diagnose issues using natural language queries
- Reporting: Generate system reports and insights
- Automation: Execute administrative tasks through conversation
[!NOTE] π Full Documentation: Agents Guide β
Additional links:
Python Clients
Simple Python client examples for testing and interacting with the MCP Server.
import asyncio
from mcp import ClientSession
from mcp.client.streamable_http import streamablehttp_client
async def main():
# Connect to the IBM i MCP server with authentication
async with streamablehttp_client("http://localhost:3010/mcp") as (
read_stream,
write_stream,
_,
):
# Create a session using the authenticated streams
async with ClientSession(read_stream, write_stream) as session:
# Initialize the connection
await session.initialize()
# List available tools (now authenticated with your IBM i credentials)
tools = await session.list_tools()
for i, tool in enumerate(tools.tools, 1):
print(f"{i:2d}. {tool.name}")
print(f" ββ {tool.description}")
# Execute a tool with authenticated IBM i access
print("\n" + "=" * 80)
print("SYSTEM ACTIVITY RESULT")
print("=" * 80)
result = await session.call_tool("system_activity", {})
print(result)
if __name__ == "__main__":
asyncio.run(main())
[!NOTE] π Full Documentation: Client README β
Additional links:
Deployment
Production-ready deployment configurations for containerized environments.
Deployment Options
- Docker & Podman - Complete stack with MCP Context Forge Gateway
- OpenShift - Kubernetes deployment with S2I builds
- Production Features - HTTPS, authentication, monitoring, caching
Setup Mapepire
Before you can use the ibmi-mcp-server, you must install and configure Mapepire on your IBM i system.
What is Mapepire?
Mapepire is a modern, high-performance database server for IBM i that provides SQL query execution capabilities over WebSocket connections. It acts as a gateway between modern application architectures (like MCP servers, AI agents, and REST APIs) and IBM i's Db2 for i database.
Why Mapepire Enables AI and MCP Workloads
Traditional IBM i database access methods (ODBC, JDBC) don't align well with modern AI and MCP architectures that require:
- Fast, lightweight connections: AI agents make frequent, short-lived database queries
- WebSocket support: Enables real-time, bidirectional communication for streaming results
- Modern JSON-based protocols: Simplifies integration with TypeScript/JavaScript ecosystems
- Low-latency responses: Essential for interactive AI conversations and tool executions
Mapepire bridges this gap by providing a modern, WebSocket-based SQL query interface that's optimized for the request/response patterns of AI agents and MCP tools.
Installation
Quick Install (IBM i SSH Session):
# 1. Install Mapepire using yum
yum install mapepire-server
# 2. Install Service Commander (if not already installed)
yum install service-commander
# 3. Start Mapepire service
sc start mapepire
[!NOTE] π Full Documentation: Mapepire System Administrator Guide β
[!IMPORTANT] Important Notes:
- By default, Mapepire runs on port
8076. You'll need this port number when configuring theDB2i_PORTvariable in your.envfile.- Ensure your IBM i firewall allows inbound connections on port 8076
- For production deployments, configure SSL/TLS certificates (see official guide)
License
This project is licensed under the Apache License 2.0. See the LICENSE file for details.
Recommended Servers
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.
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.
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.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
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.
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.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.
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.