PlugWiki
Keploy logo

Keploy MCP Server: Setup for Claude, Cursor and Other Apps

OfficialRemote
Gethttps://api.keploy.io/client/v1/mcpkeploy/keploy

Lets Claude, Cursor and other coding assistants generate API tests, run them against your API and show which endpoints have no tests.

This is the MCP endpoint built into the Keploy API server, so your assistant can call Keploy's test generation tools directly. It is hosted, and you sign in with your Keploy account or use a personal access token. The docs say agent-based test generation is available on the Keploy Enterprise plan. The server shows a few search tools first, and the assistant finds the rest by name.

Good for

  • Backend developers who want an assistant to write and run API tests for their service
  • QA engineers who want to find endpoints without tests
  • Teams that want test results synced to the Keploy platform

Works in

14 should work

Tap an app to see how to set it up and any known problems.

Install Keploy MCP

Claude Code

  1. 1

    Copy this prompt.

    Add the Keploy MCP server to Claude Code. Run this command:
    
    claude mcp add --transport http keploy https://api.keploy.io/client/v1/mcp
    
    Then check that it loaded and tell me if I need to sign in or restart.
  2. 2

    Paste it into Claude Code. It shows the change and asks before running it.

Requires Sign-in with the vendor account (OAuth).

Sign-in (OAuth) is what the generated method uses; the server answers unauthenticated calls with an OAuth challenge. Keploy's own docs show the personal access token method, and test generation needs the Enterprise plan. Do not commit the token.

Built from the package data we hold. Not yet tested by us.

Tools

  • Finds Keploy tools by keyword

    The server lists only a few tools by default. This one searches the full catalog so the assistant can find the tool it needs.

    Try: “What Keploy tools exist for test coverage?”

    Network · search_tools

  • Shows the inputs a Keploy tool needs

    Fetches the input details for tools the assistant already knows by name.

    Try: “Show me the inputs for generateTestSuites.”

    Network · get_tool_schema

  • Runs any Keploy tool by name

    Runs a tool from the catalog by name with the arguments you give it. This is how the assistant reaches every tool below.

    Try: “Run listApps.”

    Write · invoke_tool

  • Checks whether you are signed in

    Reports whether the server has your Keploy sign-in or token.

    Try: “Am I signed in to Keploy?”

    Network · get_auth_status

  • Shows setup help

    Returns setup instructions for the server.

    Try: “How do I set up Keploy here?”

    Network · get_setup_instructions

  • Lists your applications

    Lists the applications in your Keploy account.

    Try: “List my Keploy apps.”

    Network · listApps

  • Creates an application

    Creates a new application in Keploy.

    Try: “Create a Keploy app called orders-api.”

    Write · createApp

  • Generates API tests from an OpenAPI spec

    Starts AI test generation from an OpenAPI spec.

    Try: “Generate API tests from our OpenAPI spec.”

    Write · generateTestSuites

  • Generates tests and waits until they finish

    A combined tool that starts test generation and waits for it to complete.

    Try: “Generate tests for the orders API and wait for them.”

    Write · generate_and_wait

  • Runs test suites against an API

    Runs test suites against the target API you name.

    Try: “Run the test suites against our staging server.”

    Write · runTestSuites

  • Runs tests and reports failures and gaps

    A combined tool that runs tests and returns the results with failures and coverage gaps.

    Try: “Run the suites and show me what failed.”

    Write · run_and_report

  • Lists endpoints with no tests

    Returns uncovered endpoints with suggestions ranked by priority.

    Try: “Which endpoints still have no tests?”

    Network · get_coverage_gaps

From the official docs

Permissions

medium risk · 3 things to watch

Medium risk

What it can do

  • Read your data
  • Create and change things
  • Delete things
  • Read-only mode

Watch out for

  • It creates and changes things in your Keploy account

    Tools create apps and test suites and start test generation. Use a test project while you try it. keploy.io

  • It sends test traffic to the API you name

    Running suites sends real requests to the target server. Point it at staging, not production. keploy.io

  • Tokens in config files can leak

    The docs warn not to commit API keys. Use environment variables or keep the config out of git. keploy.io

Security advisories

No published security advisories for the MCP server as of Nov 1, 2026.

3 other advisories are about the wider keploy project, not this MCP server. See them on GitHub

Technical details
Turning on read-onlyNo read-only mode is documented.
Talks toapi.keploy.io. The API server you ask it to test
Sign-in and keysSign in with your Keploy account (OAuth), or send a personal access token (kep_...) in an Authorization header. Do not commit the token.

Questions

Is the Keploy MCP server official?

Yes. It is built into the Keploy API server by the Keploy team (keploy/keploy) and is listed in the official MCP Registry as io.github.keploy/mcp.

What is the Keploy MCP server URL?

It is https://api.keploy.io/client/v1/mcp. If you self-host Keploy, use your own API server address instead.

Do I need the Enterprise plan?

The Keploy docs say agent-based test generation is available on the Enterprise plan, and list a Keploy account on that plan as a prerequisite.

Do I need an API key?

Not if your client supports sign-in, which uses your Keploy account. Otherwise create a personal access token in the Keploy app settings and send it as a bearer token.

Why do I only see a few tools?

By default the server runs in tool-search mode and lists only a few search tools. The assistant finds other tools by name and calls them through the invoke tool, so none are out of reach.

How many tools does it have?

The docs say 43 or more, because every endpoint of the Keploy client API becomes a tool. This page lists the 18 the docs name, so the list is partial.

Can I use it without MCP?

Yes. The docs describe a command-line workflow for environments without MCP, where the agent runs the Keploy command and reads its output.

Is it open source?

The Keploy repository is Apache-2.0. The hosted test generation platform needs a Keploy account.

Information

updated Nov 2, 2026

Publisher
Official
Endpoint
https://api.keploy.io/client/v1/mcp
Transport
Streamable HTTP
Auth
OAuth
Version
1.0.0
License
Apache-2.0
Last commit
Oct 31, 2026
Open issues
381
Listed in
Official MCP registry

WebsiteGitHubOfficial MCP RegistryChangelogIssues

Sources

Docs, issue trackers and discussions this page is built from. Issues, quotes and findings above link to their own source.

Show all 2
  1. Keploy MCP OAuth metadata · api.keploy.io
  2. agent-test-generation · keploy.io

Facts from the official MCP registry, GitHub, npm, PyPI and the Docker catalog. Updated Nov 2, 2026.