api-testing MCP servers & tools

A MCP server for replacing Rest Clients like Postman/Insomnia, by allowing your LLM to maintain and use api collections.

The most complete API testing MCP — Postman inside your AI assistant. Per-project environment groups with {{variable}} interpolation, saved collections, multi-step flows with response extraction, assertions, load testing, and OpenAPI mock generation. 42 tools, zero cloud account needed.

An MCP server to manage test suites and cases, which is an alternative to Postman.

CLI-first API testing tool with structured JSON failures (failurecategory, errorcode, hints) so AI agents can branch on the taxonomy without parsing stderr. MCP tools: tarnrun, tarnvalidate, tarnfixplan, tarninspect, tarnrerunfailed, and more. Tests are .tarn.yaml. Single static binary (musl), MIT.

A basic MCP server to operate on the Postman API.

MCP server for James Bloom MockServer - create expectations, verify requests, and manage mock state

Deterministic API regression checks for AI agents: snapshot a baseline, catch broken contracts.

End-to-end API testing — generate and run tests from OpenAPI, curl, Postman, or real user traffic.

A MCP server for interacting with Bruno API Client.

Deterministic OpenAPI testing and policy-gated API execution for AI agents.

MCP server for Dev Proxy that enables AI assistants to interact with Dev Proxy.

MCP server for describing and invoking gRPC methods

Debug webhooks from your AI agent: inspect and replay captured webhooks on localhost.

HTTP client, JWT decode, header analysis for AI agents

API Schema Drift Monitor — detect breaking changes in OpenAPI specs

Call the real API and check the response against your OpenAPI spec.

Agent-native API client: execute/test REST/GraphQL/WS/SSE with assertions, extraction, collections

Scaffold, validate, format, and run YAML .mmt API tests. Do not invent syntax.

Python MCP server for running and inspecting Bruno API collections.