Connect Your Client
Add the MCP Client Tester server to your LLM client in minutes. This public server helps you test and debug MCP protocol implementations.
Public Server Endpoint
https://api.mcp-tester.l.supported.systems/mcpThis endpoint supports Streamable HTTP transport (JSON-RPC over HTTPS).
Quick Setup by Client
Claude.ai Web Client Setup
Claude.ai (the web interface at claude.ai) supports MCP integrations for Pro, Team, and Enterprise users.
Setup Steps:
- Go to claude.ai and sign in
- Click your profile icon → Settings
- Navigate to Integrations or MCP Servers
- Click Add Integration or Add MCP Server
- Enter:
- Name:
MCP Client Tester - URL:
https://api.mcp-tester.l.supported.systems/mcp
- Name:
- Click Save or Connect
Once connected, you can ask Claude to use the MCP tools:
Use the ping tool from MCP Client Tester with message "hello"Claude Code Setup
Add the MCP server using the CLI:
claude mcp add mcp-tester --transport http https://api.mcp-tester.l.supported.systems/mcpOr add it manually to your Claude Code settings:
Via Settings UI:
- Open Claude Code settings
- Navigate to MCP Servers
- Click Add Server
- Enter:
- Name:
mcp-tester - Transport:
http - URL:
https://api.mcp-tester.l.supported.systems/mcp
- Name:
Via config file (~/.claude/settings.json):
{ "mcpServers": { "mcp-tester": { "transport": "http", "url": "https://api.mcp-tester.l.supported.systems/mcp" } }}Claude Desktop Setup
Claude Desktop primarily uses STDIO transport. For HTTP-based servers, you’ll need a bridge or use experimental HTTP support (v1.3+).
Option 1: Using mcp-remote (Recommended)
Install the mcp-remote bridge:
npm install -g @anthropic/mcp-remoteAdd to your claude_desktop_config.json:
{ "mcpServers": { "mcp-tester": { "command": "mcp-remote", "args": [ "--transport", "http", "--url", "https://api.mcp-tester.l.supported.systems/mcp" ] } }}Option 2: Direct HTTP (v1.3+ experimental)
{ "mcpServers": { "mcp-tester": { "transport": "http", "url": "https://api.mcp-tester.l.supported.systems/mcp" } }}Python MCP SDK
from mcp import Clientfrom mcp.transports.http import HTTPClientTransport
async def connect_to_tester(): transport = HTTPClientTransport( url="https://api.mcp-tester.l.supported.systems/mcp" )
async with Client(transport) as client: # Initialize connection await client.initialize()
# List available tools tools = await client.list_tools() print(f"Available tools: {[t.name for t in tools]}")
# Call the ping tool result = await client.call_tool("ping", {"message": "hello"}) print(f"Ping response: {result}")TypeScript MCP SDK
import { Client } from '@modelcontextprotocol/sdk/client';import { HTTPClientTransport } from '@modelcontextprotocol/sdk/client/http';
const transport = new HTTPClientTransport({ url: 'https://api.mcp-tester.l.supported.systems/mcp'});
const client = new Client(transport);
await client.connect();
// List available toolsconst tools = await client.listTools();console.log('Available tools:', tools.map(t => t.name));
// Call the ping toolconst result = await client.callTool('ping', { message: 'hello' });console.log('Ping response:', result);Direct HTTP Requests
Test the MCP endpoint directly with cURL:
Initialize Connection:
curl -X POST https://api.mcp-tester.l.supported.systems/mcp \ -H "Content-Type: application/json" \ -d '{ "jsonrpc": "2.0", "id": 1, "method": "initialize", "params": { "protocolVersion": "2024-11-05", "capabilities": {}, "clientInfo": { "name": "my-test-client", "version": "1.0.0" } } }'List Tools:
curl -X POST https://api.mcp-tester.l.supported.systems/mcp \ -H "Content-Type: application/json" \ -d '{ "jsonrpc": "2.0", "id": 2, "method": "tools/list", "params": {} }'Call a Tool:
curl -X POST https://api.mcp-tester.l.supported.systems/mcp \ -H "Content-Type: application/json" \ -d '{ "jsonrpc": "2.0", "id": 3, "method": "tools/call", "params": { "name": "ping", "arguments": {"message": "hello from curl"} } }'Available Tools
Once connected, you’ll have access to these testing tools:
| Tool | Description |
|---|---|
ping | Basic connectivity test - returns pong with timestamp |
echo | Echo back any input you send |
tell_us_about_yourself | Client self-reporting for capability testing |
test_progress | Test progress notification handling |
test_cancellation | Test operation cancellation support |
test_error | Test error handling with various error types |
get_server_capabilities | Discover server features and configuration |
Testing Your Connection
After adding the server, verify the connection works:
-
Check Server Status
Use the
pingtool to verify connectivity:Use the ping tool with message "hello" -
Explore Available Tools
Ask your client to list available tools or use
get_server_capabilities -
Test Protocol Features
Try
test_progressto see how your client handles progress notifications -
Test Error Handling
Use
test_errorwith different error types to verify error handling
Troubleshooting
Connection Issues
“Connection refused” or timeout:
- Verify the URL is correct:
https://api.mcp-tester.l.supported.systems/mcp - Check your network allows HTTPS connections
- Try the cURL test above to isolate client vs. network issues
“Transport not supported”:
- Ensure your client supports HTTP/HTTPS transport
- Some clients (older Claude Desktop) only support STDIO
- Use mcp-remote bridge for STDIO-only clients
JSON-RPC errors:
- Check that your client sends proper JSON-RPC 2.0 format
- Ensure Content-Type header is
application/json - Verify the method name and parameters match the schema
Getting Help
- View Logs: Check the Monitor page for real-time protocol logs
- API Docs: Visit
https://api.mcp-tester.l.supported.systems/docsfor OpenAPI documentation - Dashboard: Use the Dashboard to see session statistics
Ready to dive deeper? Check out the Protocol Overview to understand the MCP message format, or explore Testing Strategies for comprehensive testing.