What is the MCP Debugger?
The MCP Debugger tests and diagnoses MCP (Model Context Protocol) server connections on Apify standby URLs. It sends a real protocol handshake — the same initialize request that Claude Desktop and other MCP clients send — and reports exactly what works, what fails, and how to fix it. MCP servers on Apify run as standby actors: always-on HTTP endpoints that accept JSON-RPC requests at a /mcp path. When they stop working, the error messages from MCP clients are frustratingly vague: 'connection failed', 'transport error', or just silence. The MCP Debugger replaces guesswork with diagnostics. Here is what the MCP Debugger checks. Connection success — can it reach the standby URL and get a response? Latency — how long does the initialize handshake take? Under 2 seconds is healthy, 2-5 seconds suggests a cold start, over 5 seconds is a warning. Protocol compliance — does the server respond with valid JSON-RPC? Tool discovery — how many tools does the server expose, and what are their names and descriptions? Authentication — does the server require a token, and does the provided token work? The debugger diagnoses common issues and maps them to actionable fixes. A 404 means your webServerMcpPath is wrong in actor.json. A timeout means standby mode is not enabled or the actor has not been warmed up. A 401 means the authentication token is missing or invalid. Zero tools returned means your server.tool() calls happen after the transport connects instead of before. The MCP Debugger costs $0.15 per debug run. To use it, go to the MCP Debugger page in the ApifyForge dashboard, enter your standby URL (e.g., https://username--actor-name.apify.actor/mcp), optionally provide an API token, and click Debug Connection. The report shows connection status, latency, available tools, detected issues, and fix suggestions. This tool is especially valuable when setting up a new MCP server, troubleshooting user-reported connection failures, and monitoring MCP server health over time. For related testing tools, see the questions about the Schema Validator, Test Runner, and Cloud Staging. Visit apifyforge.com/tools/mcp-debugger for documentation.
Related questions
The Schema Validator checks your actor's actual dataset output against the schema you declared in your dataset_schema.js...
What is the Test Runner?The Test Runner is a testing tool that runs your Apify actor with predefined test inputs and automatically validates the...
What is Cloud Staging?Cloud Staging runs your actor in Apify's actual production environment before you make it public on the Store, catching ...
What are Regression Tests?Regression tests are automated test suites that run before every publish to verify that new code changes have not broken...