While using an Autodesk MCP server, you may occasionally encounter situations where a request does not produce the expected result. This does not necessarily indicate a problem with the server; instead, it may reflect how the MCP server was designed to operate, or it may point to a configuration or connectivity issue.
Before you can send a request, your AI application must establish a connection to the MCP server and, if required, authenticate successfully. Common causes of failure at this stage include:
For servers that require authentication, you may also encounter failures if:
For remote MCP servers, check your network connection. If the server requires authentication, verify that your credentials are valid and up to date. For local MCP servers, verify that the server process is running and has not crashed or become unresponsive.
Your request may fail during validation, before the server executes the requested tool or returns the requested resource. This happens when the server determines that it cannot safely or correctly process your request. Common causes include:
When this happens, review your request and ensure that it aligns with the server's documented tool schemas and input requirements.
Even if you submit a well-formed request, the server may reject it based on policy rules. Common causes include:
For servers that require authentication, the server may also reject your request if you do not have permission to perform the requested action.
If the server rejects your request, consult the server's documentation to understand any access requirements or limitations.
In some cases, your request may succeed but the result may not be what you expected. Common causes include:
When this happens, refine your request or try a different approach, such as using a prompt provided by the server to guide the interaction.
When your request fails, you typically receive a structured error response that includes an error code and a descriptive message. Use these details to diagnose the issue and determine the appropriate next step.
Check the error response in your AI application's output or logs to understand why your request failed. Note that some failures, such as network timeouts or server crashes, may not return a structured error response at all.
Autodesk MCP servers execute only what you explicitly request and only within their defined boundaries. When something does not work, start by checking error messages. Then verify your inputs and, if applicable, your authentication. If the issue persists, consult server-specific documentation to resolve it.