Inspector playground
Connect local or remote MCP servers, call tools, prompts, and resources, and inspect exact arguments and results in a local playground.
MCPJam
MCPJam is a developer platform for testing, debugging, and evaluating Model Context Protocol servers across models and AI clients. It combines a local Inspector with user testing, simulated swarms, repeatable evals, and CI/CD checks for pre-production workflows.
MCP server testing and evaluation platform
MCPJam is a testing and debugging platform for Model Context Protocol (MCP) servers. It provides an open-source Inspector for connecting to local or remote servers, calling their capabilities, comparing model behavior, and examining JSON-RPC traffic and end-to-end traces.
The platform extends interactive testing into repeatable workflows: teams can run user tests, simulated swarms, cross-client evaluations, and CI/CD checks. It is designed for pre-production validation rather than live-traffic instrumentation, helping developers investigate failures and detect behavior regressions before release.
MCPJam features
Multi-model traces
Compare multiple models side by side on one prompt and read a user-to-agent-to-server waterfall trace with individual tool-call timing.
OAuth and access debugging
Visualize OAuth and EMA/XAA handshakes, with specification conformance checks to help identify where authentication flows break.
Cross-client evaluations
Generate repeatable evaluation suites from server tools, run them across selected clients, and review accuracy, latency, and diagnosed failure patterns.
Swarms and user testing
Run simulated user journeys across personas, goals, and selected clients, or share hosted user-testing environments whose sessions can become replayable regression tests.
CLI, SDK, and CI/CD checks
Run the same checks headlessly through the CLI or SDK and add MCPJam evaluation gates to GitHub Actions so pull requests can be blocked on behavior regressions.
How teams use MCPJam
Debug a local MCP server
Connect a server during development, call its tools or resources, and inspect raw arguments, results, JSON-RPC messages, and timing to reproduce implementation problems.
Check cross-client behavior
Run the same workflow across ChatGPT, Claude, and other selected clients to compare tool selection and response quality, rather than validating protocol conformance in only one environment.
Acceptance-test user journeys
Describe support, finance, or developer workflows and let simulated personas execute multi-turn journeys across chosen clients, exposing edge cases before launch.
Detect regressions in evaluation suites
Turn important workflows into durable suites, review pass rates, latency, token use, and tool-call counts, and rerun them as the server changes.
Gate releases in CI/CD
Add evaluation and conformance checks to a repository workflow so a pull request can be blocked when model behavior falls below a required threshold.
Pros and Cons
Pros
- Open-source Inspector, core CLI, and SDK can be run locally or in a user's own CI.
- Supports testing tools, prompts, resources, authentication flows, model behavior, and JSON-RPC exchanges in one workflow.
- Cross-client evaluations, simulated users, and CI/CD gates extend testing beyond manual protocol checks.
- Supports local-server testing through a built-in tunnel without requiring an initial deployment.
Cons
- Some automated workflows consume credits, with usage varying by feature; hosted-model sessions can also use credits unless users bring their own API keys.
- The detailed source material focuses on pre-production testing, so it does not establish live-traffic monitoring or production observability capabilities.
FAQ
How do I get started with MCPJam?
MCPJam can be started through the web app, desktop app, or locally with `npx @mcpjam/inspector@latest`. The open-source Inspector, core CLI, and SDK can run locally or in a user's own CI without signup.
Can I test a local MCP server in real AI clients?
Yes. MCPJam supports local and remote MCP servers. Its built-in tunnel can expose a local server to ChatGPT, Claude, or Gemini over a secure URL for testing without deploying the server first.
What MCP servers can MCPJam test?
MCPJam is described as SDK-, framework-, and deployment-agnostic. If an MCP client can connect to a server, MCPJam can test it using the same configuration shape as an `mcp.json` file.
What does MCPJam show during testing?
The Inspector tests tools, prompts, and resources and records JSON-RPC requests and responses, including direction, payload, latency, and a waterfall trace of each run.
Is MCPJam free?
The Free plan includes 200 credits per day and a one-time grant of 500 eval iterations, with no credit card required. Paid Pro, Team, and Enterprise plans add higher usage allowances and, depending on the plan, unlimited trace history, RBAC, SSO, audit logs, and enhanced support.
Quick Facts
- Category
- Developer Tool
- Primary focus
- MCP server testing, debugging, and evaluation
- Deployment model
- Web app, desktop app, local CLI, and SDK
- Supported test targets
- MCP tools, prompts, resources, OAuth flows, and model-driven behavior
- Open-source components
- Inspector, core CLI, and SDK
- Free access
- 200 credits per day plus a one-time 500 eval-iteration grant; no credit card required
Alternative a MCPJam
ByteAsk
ByteAsk is a terminal-first AI coding agent for C and C++ that edits repositories and verifies changes with the real compiler, debugger, sanitizers, and tests before showing a diff. It offers a free tier plus paid plans, with editor connectors and zero-retention handling described in the source.
CreateOS Sandbox
CreateOS Sandbox is an isolated compute environment for running code and agent workloads inside Firecracker micro-VMs. It is designed for workflows that need machine-level isolation, private networking between sandboxes, and programmatic control through SDK, CLI, or MCP.
GitBot
GitBot is a local browser-based workspace for creating reusable coding bots powered by Claude Code, Codex, or OpenCode. It helps developers run those agents across repository folders with persistent threads and configurable permissions.
Manta AI
Manta AI is an autonomous web app testing tool for teams that want to map application behavior, catch regressions, and generate tests without writing scripts or maintaining selectors. It works from a URL and supports plain-English test flows, run results with screenshots, and scheduled or deployment-triggered checks.
AakarDev AI
AakarDev AI helps teams manage AI provider access, project-level setups, logs, and analytics from one dashboard. It supports BYOK workflows and lists providers including OpenAI, Google Gemini, Anthropic, Groq, Mistral AI, and Perplexity AI.
Trigger.dev chat agent
Trigger.dev chat agent is a durable AI chat backend for developers building stateful conversations that can survive refreshes, crashes, and long-running turns. It connects with the AI SDK `useChat` flow and runs on managed infrastructure with no timeout on a turn.