A Spring Boot-based Model Context Protocol (MCP) server that exposes Hoverfly as a programmable tool for AI assistants like Cursor, Claude Desktop, GitHub Copilot, or any other assistant supporting MCP. It enables dynamic mocking of third-party APIs to unblock development and testing when external services are unavailable.
Note: The following prerequisites and build steps are only required if you want to contribute or build from source. For most users, using the Docker image is sufficient.
- Java 17
- Maven 3.6+
git clone <repository-url>
cd hoverfly-mcp-server
mvn clean package
java -jar target/hoverfly-mcp-server-<version>.jar
Replace
<version>
with the actual version of the JAR you built.
To use this server with an AI assistant that supports Model Context Protocol (MCP), add the following to your .mcp.json
or mcp.config.json
:
{
"mcpServers": {
"hoverfly-mcp": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"-p 8500:8500",
"-p 8888:8888",
"-v /path/to/your/data:/opt/hoverfly-mcp/simulation-data",
"docker.io/kapish88/hoverfly-mcp-server"
]
}
}
}
8500
: Hoverfly proxy port (mocked services)8888
: Hoverfly admin port (mock control API)-v /path/to/your/data:/opt/hoverfly-mcp/simulation-data
: Mount a volume for simulation persistence
Make sure Docker is installed. The image will be pulled automatically if not available locally.
Tool Name | Description |
---|---|
get_hoverfly_status |
Checks if Hoverfly is running |
start_hoverfly_web_server |
Starts Hoverfly in simulate mode as a web server. By default, it will auto-load the most recent simulation from /opt/hoverfly-mcp/simulation-data if available, unless auto-load is disabled. |
stop_hoverfly_server |
Stops Hoverfly and clears mocks |
fetch_hoverfly_version |
Returns Hoverfly version |
list_hoverfly_mocks |
Lists all active mock APIs (request-response pairs) |
add_hoverfly_mock |
Adds a mock API using a JSON RequestResponsePair definition |
clear_hoverfly_mocks |
Removes all existing mock APIs |
show_hoverfly_endpoints_info |
Returns key Hoverfly endpoints and example usage |
get_hoverfly_documentation |
Returns Hoverfly documentation for a specific topic |
suggest_hoverfly_matchers |
Suggests matcher options for a given request-response pair JSON |
get_hoverfly_debug_logs |
Fetches recent Hoverfly logs for debugging (limit is optional) |
download_hoverfly_simulation |
Downloads current simulation to /opt/hoverfly-mcp/simulation-data (persistent simulation directory) |
These tools can be invoked programmatically by AI assistants through the AI Assistant host.
The server supports simulation persistence through a fixed, mounted volume:
When starting Hoverfly with start_hoverfly_web_server
, the server will by default:
- Load the most recent simulation file from
/opt/hoverfly-mcp/simulation-data
(if available) - Start with a clean state if no simulation file is found
You can disable auto-load by setting the appropriate parameter.
To persist simulation data across restarts, you must mount a host directory to /opt/hoverfly-mcp/simulation-data
inside the container.
This is the only supported location for persistent data.
Note: The host directory must be writable by the container user.
Simulation files are only saved to this directory when you explicitly invoke the download tool.
- Fork this repo
- Create a feature branch
- Submit a Pull Request
See LICENSE file for licensing terms.