hlf-fabric-agent

Hyperledger Fabric Agent Suite

https://github.com/user-attachments/assets/75c3414a-3901-42ff-bf33-8f57b0ca96b4

Works for test-network.

A modular toolkit for managing Hyperledger Fabric test networks and chaincode lifecycle, featuring:

• hlf-controller: A Go-based REST API for network and chaincode operations.
• hlf-mcp: A Python-based MCP tool for automation, scripting, and agent integration.

---

Features

• Automated Fabric Setup: The hlf-controller automatically downloads and configures Hyperledger Fabric binaries and samples on first run.
• Dynamic Configuration: Easily configure the entire Fabric network topology via a central config.yaml file.
• Full Lifecycle Management: Bring up/down the network, create channels, deploy, invoke, and query chaincode via a simple REST API.
• Agent-Ready: The hlf-mcp tool allows LLMs (like those in Cursor or Claude Desktop) to interact with your Fabric network seamlessly.

---

Prerequisites

• Go (1.18+)
• Python (3.9+)
• pip (for Python dependencies)

---

Setup Instructions

1. Clone the Repo

git clone <your-repo-url>
cd <repo-root>

2. Set Up and Run hlf-controller (Go REST API)

The controller handles its own prerequisites.

cd hlf-controller
go run main.go

• The first time you run this, it will check for fabric-samples and download them to your home directory if they are missing. The server will not start until this process is complete.
• The API will then be available at http://localhost:8081.

3. Set Up hlf-mcp (Python MCP Tool)

cd ../hlf-mcp
pip install httpx
# (Optional) Install any other agent/MCP dependencies

---

Configuration

hlf-controller

The Go controller is configured via the hlf-controller/config.yaml file. This is the primary way to define your network topology, including peer/orderer details, certificate paths, and timeouts.

Environment Variables

• HLF_CONFIG_PATH: Path to a custom config.yaml file for the controller.
• HLF_NETWORK_SCRIPT_PATH: Overrides the path to your Fabric network.sh. If not set, the path from config.yaml is used, which in turn defaults to ~/fabric-samples/test-network/network.sh.
• HLF_API_BASE: Base URL for the hlf-controller API (used by hlf-mcp, defaults to http://localhost:8081).

---

How to Use

Integration

• Cursor: Add the hlf-mcp tool to your ~/.cursor/mcp.json (see hlf-mcp/README.md).
• Claude Desktop: Add to claude_desktop_config.json (see hlf-mcp/README.md).

---

Example JSON Configurations for Integration

Cursor (~/.cursor/mcp.json)

Add the following entry to your ~/.cursor/mcp.json file to integrate the MCP tool with Cursor:

"hlf-controller": {
    "command": "uv",
    "args": [
        "--directory",
        "/Users/padamarajkore/Desktop/hlf-mcp", //change it to your local path for hlf-mcp folder
        "run",
        "mcp_hlf_tool.py"
    ]
}

Claude Desktop (claude_desktop_config.json)

Add the following entry to your claude_desktop_config.json file (usually found in ~/Library/Application Support/Claude/claude_desktop_config.json on macOS):

"hlf-controller": {
    "command": "/Users/padamarajkore/.local/bin/uv", //change it to your local path for uv binary
    "args": [
        "--directory",
        "/Users/padamarajkore/Desktop/hlf-mcp", //change it to your local path for hlf-mcp folder
        "run",
        "mcp_hlf_tool.py"
    ]
}

Note:

• Adjust the --directory and script name to match the actual location and filename of your MCP tool.
• Adjust the path to the uv command if it is installed elsewhere on your system.
• Place these entries inside the top-level JSON object, alongside your other tool definitions.

---

Project Structure

.
├── hlf-controller/        # Go REST API server with its own config.yaml
├── hlf-mcp/               # Python MCP tool
├── README.md              # This file

---

License

MIT