Mifos MCP - Model Context Protocol (MCP)

This project provides Model Context Protocol (MCP) for the Mifos X Ecosystem, enabling AI agents to access financial data and operations.

Implementations are available in:

• Go (Native) — 102 typed tools (high-performance, cloud-native with SSE/Stdio).
• Java (Quarkus) — 38 typed tools (across Backoffice and Recommendations).
• Python (FastMCP) — 49 typed tools (modular domain-driven design).
• Rust — 89 typed tools (high-performance async I/O with exclusive bulk operations).

The Mifos MCP Server acts as a standalone, stateless integration tier that bridges any AI assistant or agent framework to the Apache Fineract banking backend.

┌──────────────────────────────────────────────┐
│            Apache Fineract / Mifos X          │
└───────────────────────┬──────────────────────┘
                        │ REST API
┌───────────────────────────────▼───────────────────────────────┐
│                 mcp-mifosx (Primary Repo)                     │
│                                                               │
│  ┌─────────────────┐ ┌─────────────────┐ ┌─────────────────┐ ┌─────────────────┐  │
│  │   /go (Native)  │ │ /java (Quarkus) │ │/python (FastMCP)│ │  /rust (Tokio)  │  │
│  │                 │ │                 │ │                 │ │                 │  │
│  │ - 102 Tools     │ │ - 38 Tools      │ │ - 49 Tools      │ │ - 89 Tools      │  │
│  │ - Go Routines   │ │ - Backoffice    │ │ - Modular Design│ │ - Async I/O     │  │
│  │ - SSE / Stdio   │ │ - Recommend.    │ │                 │ │ - Bulk Actions  │  │
│  └────────┬────────┘ └────────┬────────┘ └────────┬────────┘ └────────┬────────┘  │
└───────────┴───────────────────┼───────────────────┼───────────────────┴───────────┘
                        │ MCP Protocol (stdio / SSE)
          ┌─────────────┼──────────────┐
          ▼             ▼              ▼
    Mifos X WebApp   Claude Code     n8n / Custom
    AI Assistant     (claude.ai)     Workflow Agent
    (your client)   (external)       (your client)

This repository is framework-agnostic. The client (LLM brain, UI, memory) lives in a separate repository. Any MCP-compatible system can plug in.

While this repository hosts two different programming languages, they are kept in functional parity where possible to ensure a consistent experience.

1. Tool Specification: All implementations aim to expose the same core banking tools.
   • Go currently leads with 102 tools, featuring advanced cloud-native features and SSE.
   • Rust provides 89 tools, uniquely featuring high-concurrency Bulk Operations and robust "Fetch-and-Merge" state management.
   • Python provides 49 tools using a modular domain design.
   • Java provides 38 tools (21 for Backoffice operations and 17 for User Recommendations).
2. API Alignment: All implementations are built against the same Apache Fineract REST API. They share identical logic for field routing.
3. Stateless Parity: All implementations follow a strictly stateless design. None of the servers store user data, PII, or AI memory.
4. Testing Protocol: Shared "Smoke Tests" ensure that all implementations return identical, predictable JSON structures to the LLM.

This repository is structured to support multiple implementations and client integrations.

.
├── README.md               # Root entry point & cross-implementation guide
├── go/                     # Go Implementation (Native / High-Performance)
│   ├── tools/              # 102 Domain-specific tools (SSE/Stdio)
│   ├── server/             # Dual-transport logic (HTTP/SSE & Stdio)
│   └── main.go             # Server entry point
├── rust/                   # Rust Implementation (Tokio/Reqwest)
│   ├── src/                # Multi-threaded typed tools & bulk execution logic
│   └── Cargo.toml          # Rust package dependencies
├── python/                 # Python Implementation (FastMCP)
│   ├── mcp_server.py       # Main entry point for the MCP server
│   ├── tools/              # Domain-specific banking tools (Loans, Clients, etc.)
│   └── core/               # API Gateway handlers
└── java/                   # Java Implementation (Quarkus)
    ├── backoffice/         # Core banking tools
    └── userrecommendation/ # Recommendation engine tools

Prerequisites: Go 1.21+

Steps:

1. Configure Environment: Copy go/.env.example to go/.env and update credentials.
2. Build and Run:

   cd go
   go build -o mcp-server .
   ./mcp-server

3. SSE Mode (Optional): Define PORT=8080 in .env to switch from Stdio to SSE microservice mode.

Prerequisites: Rust (Cargo)

Steps:

1. Configure Environment: Copy rust/.env.example to rust/.env and update credentials.
2. Build and Run:

   cd rust
   cargo build --release
   ./target/release/mcp-rust-mifosx

Prerequisites: JDK 21+, Maven

Steps:

1. Configure Environment Variables:

   export MIFOSX_BASE_URL="https://your-fineract-instance"
   export MIFOSX_BASIC_AUTH_TOKEN="your_api_token"
   export MIFOS_TENANT_ID="default"

2. Run via JBang:

   jbang --quiet org.mifos.community.ai:mcp-server:1.0.0-SNAPSHOT:runner

3. Build Native Executable (Optional):

   cd java/backoffice
   ./mvnw package -Dnative
   ./target/mcp-server-1.0.0-SNAPSHOT-runner

Prerequisites: Python 3.10+, pip

Steps:

1. Navigate to the Python directory:

   cd python

2. Install dependencies:

   pip install -r requirements.txt

3. Configure Environment: Copy .env.example to .env and fill in your details.
4. Run the Server:

   python mcp_server.py

The exact number and categorization of tools depend on the core server implementation deployed:

The most feature-complete implementation with native concurrent routines.

• Clients & Identities: 16 Tools
• Documents & Reports: 26 Tools
• Loans & Savings: 23 Tools
• Groups & Centers: 13 Tools
• Bulk & Composite: 19 Tools (Cloud-Native)
• Accounting & Stats: 5 Tools

Built for asynchronous scale, bulk processing, and robust state-aware updates.

• Clients & Collaterals: 25 Tools
• Loans & Collaterals: 19 Tools
• Groups, Savings & Centers: 23 Tools
• Staff, Accounting & Charges: 11 Tools
• Bulk Operations: 11 Tools (Exclusive to Rust)

Domain-driven design bridging AI directly to Fineract.

• Clients & Groups: 16 Tools
• Loans & Savings: 20 Tools
• Staff & Accounting: 13 Tools

Enterprise suite categorized between Backoffice and recommendation engines.

• Backoffice Operations: 21 Tools (Covers Clients, Loans, Savings)
• User Recommendations: 17 Tools (Exclusive to Java)

Use the MCP Inspector to test and debug your server interactively:

npx @modelcontextprotocol/inspector <command_to_run_yours_server>

For Python:

npx @modelcontextprotocol/inspector python python/mcp_server.py

• Universal Compatibility — Works with Claude, GPT-4, Qwen, or any MCP client.
• Data Sovereignty — The server makes no external calls.
• RBAC Enforced — Every action is validated against Fineract's native permissions.

• Mifos Community: https://mifos.org
• Mifos MCP (Docker): https://hub.docker.com/r/openmf/mifos-mcp
• Chatbot Demo: https://ai.mifos.community