Home / MCP / Postman MCP Server
Runs Postman collections via Newman and exposes detailed test results for MCP-based API testing.
Configuration
View docs{
"mcpServers": {
"postman_runner": {
"command": "node",
"args": [
"/absolute/path/to/mcp-postman/build/index.js"
]
}
}
}You can run Postman collections through an MCP server that uses Newman under the hood, enabling automated API tests to be executed by language models and returning structured results for easy interpretation and debugging.
Use the Postman MCP Server with your MCP client to run Postman collections and retrieve detailed test results. You can point the server to a collection, optional environment and globals files, and run multiple iterations. The server returns a clear summary of success, a test-by-test breakdown, and timing information to help you spot issues quickly.
Prerequisites: ensure you have Node.js installed on your system. You may also need to install a package manager if you plan to build from source.
Install via Smithery (automatic for Claude Desktop):
npx -y @smithery/cli install mcp-postman --client claudeManual installation steps are shown below. Follow these to clone, install dependencies, and build the project.
# Clone the repository
git clone <repository-url>
cd mcp-postman
# Install dependencies
pnpm install
# Build the project
pnpm buildAfter building, you will run the server locally using Node to execute the built entry point.
node /absolute/path/to/mcp-postman/build/index.jsConfigure your Claude desktop to load the Postman MCP Server by adding an MCP server entry that points to the built index. The following example shows how to reference the local build with a stdio command.
{
"mcpServers": {
"postman_runner": {
"command": "node",
"args": ["/absolute/path/to/mcp-postman/build/index.js"]
}
}
}The server exposes a tool named run-collection that executes a Postman collection and returns a structured result set.
Parameters shown in usage outline include the collection path, optional environment and globals files, and an optional iteration count.
Example response format you can expect from a run-collection call shows overall success, a summary with total/passed/failed counts, detailed failures, and execution timings.
Project structure includes an entry point, an MCP server implementation, a Newman runner, and tests. You can run tests and build locally to verify changes.
pnpm test
pnpm test:coverage
pnpm build
pnpm cleanRuns a Postman collection, optionally with environment and globals files, and returns a structured test results payload including success status, summary, failures, and timings.