Unbundle OpenAPI MCP Server
A server for splitting and extracting parts of OpenAPI specifications using Redocly CLI.
Unbundle OpenAPI MCP Server
This project provides a Model Context Protocol (MCP) server with tools to split OpenAPI specification files into multiple files or extract specific endpoints into a new file. It allows an MCP client (like an AI assistant) to manipulate OpenAPI specifications programmatically.
Prerequisites
- Node.js (LTS version recommended, e.g., v18 or v20)
- npm (comes with Node.js)
Usage
Installing via Smithery
To install Unbundle OpenAPI MCP Server for Claude Desktop automatically via Smithery:
npx -y @smithery/cli install @auto-browse/unbundle_openapi_mcp --client claude
The easiest way to use this server is via npx, which ensures you are always using the latest version without needing a global installation.
npx @auto-browse/unbundle-openapi-mcp@latest
Alternatively, you can install it globally (not generally recommended):
npm install -g @auto-browse/unbundle-openapi-mcp
# Then run using: unbundle-openapi-mcp
The server will start and listen for MCP requests on standard input/output (stdio).
Client Configuration
To use this server with MCP clients like VS Code, Cline, Cursor, or Claude Desktop, add its configuration to the respective settings file. The recommended approach uses npx.
VS Code / Cline / Cursor
Add the following to your User settings.json (accessible via Ctrl+Shift+P > Preferences: Open User Settings (JSON)) or to a .vscode/mcp.json file in your workspace root.
// In settings.json:
"mcp.servers": {
"unbundle_openapi": { // You can choose any key name
"command": "npx",
"args": [
"@auto-browse/unbundle-openapi-mcp@latest"
]
}
// ... other servers can be added here
},
// Or in .vscode/mcp.json (omit the top-level "mcp.servers"):
{
"unbundle_openapi": { // You can choose any key name
"command": "npx",
"args": [
"@auto-browse/unbundle-openapi-mcp@latest"
]
}
// ... other servers can be added here
}
Claude Desktop
Add the following to your claude_desktop_config.json file.
{
"mcpServers": {
"unbundle_openapi": {
// You can choose any key name
"command": "npx",
"args": ["@auto-browse/unbundle-openapi-mcp@latest"]
}
// ... other servers can be added here
}
}
After adding the configuration, restart your client application for the changes to take effect.
MCP Tools Provided
split_openapi
Description: Executes the redocly split command to unbundle an OpenAPI definition file into multiple smaller files based on its structure.
Arguments:
apiPath(string, required): The absolute path to the input OpenAPI definition file (e.g.,openapi.yaml).outputDir(string, required): The absolute path to the directory where the split output files should be saved. This directory will be created if it doesn't exist.
Returns:
- On success: A text message containing the standard output from the
redocly splitcommand (usually a confirmation message). - On failure: An error message containing the standard error or exception details from the command execution, marked with
isError: true.
Example Usage (Conceptual MCP Request):
{
"tool_name": "split_openapi",
"arguments": {
"apiPath": "/path/to/your/openapi.yaml",
"outputDir": "/path/to/output/directory"
}
}
extract_openapi_endpoints
Description: Extracts specific endpoints from a large OpenAPI definition file and creates a new, smaller OpenAPI file containing only those endpoints and their referenced components. It achieves this by splitting the original file, modifying the structure to keep only specified paths, and then bundling the result.
Arguments:
inputApiPath(string, required): The absolute path to the large input OpenAPI definition file.endpointsToKeep(array of strings, required): A list of the exact endpoint paths (strings) to include in the final output (e.g.,["/api", "/api/projects/{id}{.format}"]). Paths not found in the original spec will be ignored.outputApiPath(string, required): The absolute path where the final, smaller bundled OpenAPI file should be saved. The directory will be created if it doesn't exist.
Returns:
- On success: A text message indicating the path of the created file and the standard output from the
redocly bundlecommand. - On failure: An error message containing details about the step that failed (split, modify, bundle), marked with
isError: true.
Example Usage (Conceptual MCP Request):
{
"tool_name": "extract_openapi_endpoints",
"arguments": {
"inputApiPath": "/path/to/large-openapi.yaml",
"endpointsToKeep": ["/users", "/users/{userId}/profile"],
"outputApiPath": "/path/to/extracted-openapi.yaml"
}
}
Note: This server uses npx @redocly/cli@latest internally to execute the underlying split and bundle commands. An internet connection might be required for npx to fetch @redocly/cli if it's not cached. Temporary files are created during the extract_openapi_endpoints process and automatically cleaned up.
Development
If you want to contribute or run the server from source :
- Clone: Clone this repository.
- Navigate:
cd unbundle_openapi_mcp - Install Dependencies:
npm install - Build:
npm run build(compiles TypeScript todist/) - Run:
npm start(starts the server using the compiled code indist/)
Related Servers
Scout Monitoring MCP
sponsorPut performance and error data directly in the hands of your AI assistant.
Alpha Vantage MCP Server
sponsorAccess financial market data: realtime & historical stock, ETF, options, forex, crypto, commodities, fundamentals, technical indicators, & more
BlenderMCP
Connects Blender to Claude AI via the Model Context Protocol (MCP), enabling direct interaction and control for prompt-assisted 3D modeling, scene creation, and manipulation.
MCP-ABI
Interact with Ethereum-compatible smart contracts using their ABI.
MCP Yeoman Server
Search for and run Yeoman generator templates programmatically.
FLUX Image Generator
Generate images using Black Forest Lab's FLUX model.
GhostMCP
Injectable MCP server for AI-driven reverse engineering inside processes
VSCode MCP
Enables AI agents and assistants to interact with VSCode through the Model Context Protocol.
MemGPT MCP Server
A server that provides a memory system for LLMs, enabling persistent conversations with various providers like OpenAI, Anthropic, and OpenRouter.
Choose MCP Server
An MCP server for integration with the Claude Desktop Client, with optional DBT manifest path configuration.
Brainfaq
MCP server for the Brainfuck programming language that allows your favourite LLM to debug Brainfuck programs.
Sugar
Autonomous AI development system for Claude Code with task queue management and workflow automation.