Skip to main content
The MCP Inspector is an essential debugging tool that lets you test and troubleshoot your MCP servers without a full AI host application. Think of it as “Postman for MCP” — it provides a visual interface to send requests, view responses, and understand how your server behaves.

​
Why use the MCP Inspector?

When building MCP servers, you’ll commonly ask:
  • “Is my server even running?” — Inspector shows connection status
  • “Are my tools registered correctly?” — Inspector lists all available tools
  • “What’s the response format?” — Inspector displays full JSON responses
  • “Why isn’t this tool working?” — Inspector shows detailed error messages

​
Prerequisites

​
Installation

​
Connecting to your server

​
stdio servers (local process)

Pass the server startup command directly to the Inspector: 
# Python server
npx @modelcontextprotocol/inspector python server.py

# Node.js/TypeScript server
npx @modelcontextprotocol/inspector node ./build/index.js

# .NET server
npx @modelcontextprotocol/inspector dotnet run

# With environment variables
OPENAI_API_KEY=xxx npx @modelcontextprotocol/inspector python server.py

​
SSE/HTTP servers (remote)

Start your server first, then connect the Inspector to it: 
# Start the server in terminal 1
python server.py  # running on http://localhost:8080

# Launch Inspector and point to SSE endpoint in terminal 2
npx @modelcontextprotocol/inspector --sse http://localhost:8080/sse

​
Inspector interface

The web interface (typically at http://localhost:5173) shows three tabs: 
┌─────────────────────────────────────────────────────────────┐
│  MCP Inspector                              [Connected ✅]   │
├─────────────────────────────────────────────────────────────┤
│                                                             │
│  ┌─────────────┐  ┌─────────────┐  ┌─────────────┐         │
│  │   🔧 Tools  │  │ 📄 Resources│  │ 💬 Prompts  │         │
│  │    (3)      │  │    (2)      │  │    (1)      │         │
│  └─────────────┘  └─────────────┘  └─────────────┘         │
│                                                             │
│  ┌───────────────────────────────────────────────────────┐ │
│  │  📋 Message Log                                       │ │
│  │  → initialize                                         │ │
│  │  ← initialized (server info)                          │ │
│  │  → tools/list                                         │ │
│  │  ← tools (3 tools)                                    │ │
│  └───────────────────────────────────────────────────────┘ │
└─────────────────────────────────────────────────────────────┘

​
Testing tools

1

Open the Tools tab

Inspector automatically calls tools/list. You’ll see each registered tool with its name, description, and input schema.
2

Select a tool and fill in parameters

Click a tool, fill in the form fields that correspond to its input schema.
3

Click Run Tool and view the response

Example for a calculator’s add tool:
Tool: add
Parameters:
  a: 25
  b: 17

Response:
{
  "content": [
    { "type": "text", "text": "42" }
  ]
}

​
Debugging tool errors

When a tool fails, Inspector shows structured error output: 
{
  "error": {
    "code": -32602,
    "message": "Invalid params: 'b' is required"
  }
}

​
JSON-RPC error codes

CodeMeaning
-32700Parse error — invalid JSON received
-32600Invalid request structure
-32601Method not found
-32602Invalid params
-32603Internal server error

​
Testing resources and prompts

  1. Click the Resources tab.
  2. Inspector calls resources/list — you’ll see all resource URIs, names, and MIME types.
  3. Select a resource and click Read Resource to see its content.
For parameterized resources like greeting://{name}, enter the URI with the parameter filled in.
  1. Click the Prompts tab.
  2. Inspector calls prompts/list — you’ll see all registered prompt templates.
  3. Select a prompt, fill in arguments, and click Get Prompt to see the rendered message template.

​
CLI mode

For automation and CI/CD use, add --cli to run without a browser: 
# List all tools
npx @modelcontextprotocol/inspector --cli node build/index.js --method tools/list

# Call a specific tool
npx @modelcontextprotocol/inspector --cli node build/index.js \
  --method tools/call --tool-name add --tool-arg a=5 b=3

# List resources
npx @modelcontextprotocol/inspector --cli python server.py --method resources/list
CLI mode is ideal for smoke testing in CI/CD pipelines. Combine it with tools like jq to parse the JSON output and assert expected responses.

​
Key takeaways

  • Launch the Inspector by passing your server startup command: npx @modelcontextprotocol/inspector node build/index.js
  • For stdio servers, the Inspector manages the server process. For HTTP servers, start the server first then connect with --sse.
  • Use the Tools, Resources, and Prompts tabs to interactively test each capability.
  • Add --cli for CI-friendly non-interactive testing.
  • Error codes follow the JSON-RPC 2.0 standard (-32600 to -32603).

Build docs developers (and LLMs) love

Get started for freeTalk to us