MCPJam Inspector runs three ways: a hosted web app, a desktop app for Mac and Windows, or via your terminal. The web app is HTTPS-only and has no install. Terminal and Desktop support HTTP/S and local STDIO servers.
Requirements
Node.js 20+ is only required for the terminal install (npx). The hosted web app and desktop apps have no local runtime requirements.
Hosted Web App
Open app.mcpjam.com in your browser. No install required. Always on the latest version, and you can share MCP server links with teammates the same way you'd share a Google Doc.
HTTPS MCP server URLs only (for HTTP or local STDIO servers, use Desktop or Terminal).
No STDIO, tunneling, skills, or tasks. Those require the local inspector.
Run the inspector via npx (supports HTTP/S and local STDIO):
npx @mcpjam/inspector@latest
After it starts, open the printed localhost URL in your browser.
Docker
Run MCPJam Inspector using Docker, bound to localhost for security:
docker run -p 127.0.0.1:6274:6274 mcpjam/mcp-inspector
The app is available at http://127.0.0.1:6274. Always use -p 127.0.0.1:6274:6274 (not -p 6274:6274) to keep the inspector local-only. On macOS/Windows, connect to host MCP servers via http://host.docker.internal:PORT instead of 127.0.0.1.
Key features
| Capability | Description |
| -------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ |
| App Builder | Debug your server against a model: tool calls or in-panel chat, with Chat, Trace, and Raw. OpenAI Apps SDK and MCP app UIs, text tools, Chrome DevTools-style widget emulator. Read more |
| Chat | Multi-server chat on frontier models (free). Chat, Trace, Raw; compare up to 3 models. Read more |
| OAuth Debugger | Guided MCP OAuth conformance checks: protocol versions 03-26, 06-18, 11-25; DCR, client pre-registration, CIMD. Read more |
| MCP Server Debugging | Manually run tools, resources, templates, and elicitation; full JSON-RPC logs. |
| Skills | Skills in Chat and App Builder; local filesystem only. Read more |
| Workspaces | Shared server groups with real-time team sync. Read more |
| Evals | Test cases with expected tool calls, run across LLMs, metrics. Read more |
| CLI | Run MCPJam checks, probes, and evals from the terminal. Perfect for local dev loops and CI. Read more |
| SDK | Programmatic access to MCPJam for custom tooling, scripting, and integrations. Read more |
| CI/CD | Run MCPJam checks and evals in GitHub Actions and other CI systems to gate PRs on regressions. Read more |
App Builder
Debug your server against a model using tool calls or in-panel chat, with Chat, Trace, and Raw views. Supports OpenAI Apps SDK and MCP app UIs, text tools, and a Chrome DevTools-style widget emulator to iterate on widgets locally.
Manually invoke a tool to instantly view the widget, or chat with your server using an LLM.
View all JSON-RPC messages and window.openai messages in the logs.
Change emulator device to Desktop, Tablet, or Mobile views.
Test your app's locale change, CSP permissions, light / dark mode, hover & touch, and safe area insets.
Trace view: every tool call, agent step, and JSON-RPC message.
Chat
Multi-server chat on frontier models for free, or bring your own API key. Chat, Trace, and Raw views; compare up to 3 models side-by-side. View your server's token usage.
OAuth Debugger
Guided MCP OAuth conformance checks with step-by-step explanations. Test against every version of the OAuth spec (03-26, 06-18, and 11-25). Support for client pre-registration, Dynamic Client Registration (DCR), and Client ID Metadata Documents (CIMD).
MCP Server Debugging
MCPJam contains all of the tooling to test your MCP server. Manually run tools, resources, resource templates, prompts, and elicitation flows, with full JSON-RPC observability. MCPJam has all features from the original inspector and more.
Skills
Use Skills in Chat and App Builder to extend models with local, reusable behaviors. Local filesystem only. Your data never leaves your machine. Read more
Workspaces
Group your servers into shared workspaces with real-time team sync, so everyone on your team is testing against the same configuration. Read more
Evals
Define test cases with expected tool calls and run them across multiple LLMs. Track accuracy metrics over time to catch regressions early and improve your server with every iteration. [Read more](https://docs.mcpjam.com/i
The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.
A Claude Code plugin that automatically captures everything Claude does during your coding sessions, compresses it with AI (using Claude's agent-sdk), and injects relevant context back into future sessions.
