rohitg00/ai-engineering-from-scratch · error · ValueError

unsupported input request

Error message

unsupported input request: {method}

What it means

Raised in the client-side fulfillment helper when the server requests a client capability method other than the known roots/list, sampling/createMessage, and elicitation/create. It is a generic dispatch guard: the incoming request method string is the faulty input and does not match any handler this demo client implements.

Solutions

1. Add a handler branch for the requested method (e.g. ping or another client capability) in the fulfill dispatch
2. Upgrade or align the server so it only requests capabilities the client advertises
3. Return a JSON-RPC method-not-found error to the server instead of crashing when an optional capability is unsupported

When it happens

Trigger: Thrown at certifications/claude/lessons/11-mcp-server-design-and-integration/code/main.py:590 when the library encounters an invalid state.

Common situations: See trigger scenarios.

AI-assisted analysis of rohitg00/ai-engineering-from-scratch@39ea8a1c6d (2026-08-26). Data as JSON: /api/errors/881521e3831ee297. Report an issue: GitHub.

Appendix: source

Thrown at certifications/claude/lessons/11-mcp-server-design-and-integration/code/main.py:590

            method = request.get("method")
            if method == "roots/list":
                responses[key] = {
                    "roots": [{"uri": "file:///workspace", "name": "Workspace"}],
                }
            elif method == "sampling/createMessage":
                responses[key] = {
                    "role": "assistant",
                    "content": {"type": "text", "text": "Check correctness and rollback risk."},
                    "model": "study-model",
                    "stopReason": "endTurn",
                }
            elif method == "elicitation/create":
                responses[key] = {
                    "action": "accept",
                    "content": {"goal": "find correctness risks"},
                }
            else:
                raise ValueError(f"unsupported input request: {method}")
        return responses


def streamable_http_profile() -> dict[str, Any]:
    return {
        "endpoint": "/mcp",
        "method": "POST",
        "oneMessagePerPost": True,
        "protocolSessions": False,
        "responseTypes": ["application/json", "text/event-stream"],
        "requestScopedSSE": True,
        "changeNotifications": "subscriptions/listen",
        "supportsGetStream": False,
        "supportsDeleteSession": False,
        "supportsLastEventId": False,
    }

View on GitHub (pinned to 39ea8a1c6d)