

Choose your tool
MCPJam comes as three tools: Inspector, an interactive app for testing and debugging servers by hand; the CLI, for your shell and CI; and the SDK, for your test suite.Inspector
Open the Inspector in your browser, from your terminal, as a desktop app, or in Docker. The web app needs no install but only reaches HTTPS servers; the other options also run local STDIO and HTTP servers.Start with Inspector
New to MCPJam? Open the Inspector, connect a server, and send your first prompt.
Web
Runs at app.mcpjam.com. No install. HTTPS servers only. Share servers with your team.
Terminal
Requires Node.js. HTTP/S and local STDIO servers.
macOS
Desktop app (
.dmg). HTTP/S and local STDIO servers. No Node.js needed.Windows
Desktop app (
.exe installer). HTTP/S and local STDIO servers. No Node.js needed.Docker
Build the image from source. Reach host servers through
host.docker.internal.CLI and SDK
CLI
In your shell or CI. Connectivity, OAuth, conformance, and tool runs with JUnit/JSON output. Tests local servers over STDIO (
--command) and HTTP (--url).SDK
In your test suite. Unit tests, e2e tests, and evals in TypeScript with Jest or Vitest. Spawns local STDIO servers or connects over HTTP.
What you can build
Debug an MCP server end-to-end
Debug an MCP server end-to-end
Connect any HTTP, HTTPS, or STDIO server, watch every JSON-RPC message live in the Playground, and inspect tool input/output, widget state, and CSP activity per call.Open the Playground → · JSON-RPC logger →
Build an MCP App or ChatGPT App widget
Build an MCP App or ChatGPT App widget
The unified widget renderer handles both
_meta.ui.resourceUri (MCP Apps) and openai/outputTemplate (ChatGPT Apps). The window.openai shim is always injected, so you can iterate without a real host.First MCP App → · First ChatGPT App in React →Validate OAuth conformance across protocol versions
Validate OAuth conformance across protocol versions
Step through every stage of the MCP authorization flow against protocol versions 03-26, 06-18, 11-25, and 2026-07-28 (Latest), with DCR, pre-registration, and CIMD coverage.Guided OAuth → · CLI OAuth conformance →
Run evals across LLMs and gate CI on regressions
Run evals across LLMs and gate CI on regressions
Author cases with expected tool calls, optional expected outputs, and deterministic checks. Run across Claude, GPT, Gemini, and your own providers — or move the same suite into Jest/Vitest with
@mcpjam/sdk.Evaluate in Inspector → · Run evals from the SDK → · CI integration →Auto-launch Inspector from your dev workflow
Auto-launch Inspector from your dev workflow
Launch the Inspector with a prefilled server URL, bearer token, and starting tab via CLI flags — wire it into your dev script or a framework integration.Launch from code →
Capabilities at a glance
Get help
Discord
Ask the team and the community.
GitHub
File an issue or open a PR.
Changelog
What shipped, and when.

