MCPJam is a testing and evaluation platform for developers who build MCP servers. Clients such as ChatGPT, Claude, and Cursor each read a server differently, so MCPJam runs your server's tools, prompts, resources, and authorization against 16 client configurations and more than 170 models, then shows where it breaks.
You debug a server by hand or drive it with an LLM in the Playground, where a trace view puts every tool call, agent step, and JSON-RPC message on one timeline. Evals turn expected tool calls into test cases, run them across several LLMs, and track accuracy, latency, and tool-call performance over time. The same checks run from a CLI, an SDK, or a CI/CD job on every pull request.
MCPJam Inspector runs as a hosted web app (HTTPS servers only), a desktop app for Mac and Windows, or a local process from the terminal, and the desktop and terminal versions also reach local STDIO servers. It covers every feature of the original MCP Inspector.
Features
- Playground: cross-client chat that emulates client UIs, tool calls, and skills with full traces, including OpenAI Apps SDK and MCP app UIs and a widget emulator for desktop, tablet, and mobile
- OAuth Debugger: guided MCP OAuth conformance checks across spec versions 2025-03-26, 2025-06-18, 2025-11-25, and the 2026-07-28 draft, with DCR, client pre-registration, and CIMD
- Server debugging: run tools, resources, resource templates, prompts, and elicitation flows with full JSON-RPC logs
- Cross-client evals: score agent behavior across 16 client configurations, including ChatGPT, Claude, Cursor, and Copilot
- Evals: test cases with expected tool calls, run across multiple LLMs, with accuracy tracked over time
- Skills: extend models with reusable behaviors in the Playground, read locally or carried as hosted skills by a project
- Workspaces: shared server groups with real-time team sync so a team tests one configuration
- CLI and SDK: probe servers, run doctor and OAuth checks, snapshot capabilities, and assert on tool and resource shapes from your own tests
- CI/CD: run conformance, end-to-end tests, evals, and OAuth checks on every pull request in GitHub Actions, GitLab CI, or another CI system
