danielmiessler/Fabric · error · ChatError

HTTP_ERROR

HTTP_ERROR

Error message

HTTP error! status: ${response.status}

What it means

ChatService.sendMessage throws ChatError('HTTP error! status: N', 'HTTP_ERROR') when POST /api/chat returns non-2xx before streaming begins. The logged 'Final ChatRequest payload' immediately above is the exact body that was rejected, which is the key debugging artifact.

Source

Thrown at web/src/lib/services/ChatService.ts:70

				language: get(languageStore),
				pattern: get(selectedPatternName),
				promptCount: request.prompts?.length,
				messageCount: request.messages?.length,
			});
			// NEW: Log the full payload before sending to backend
			console.log(
				"Final ChatRequest payload:",
				JSON.stringify(request, null, 2),
			);

			const response = await fetch("/api/chat", {
				method: "POST",
				headers: { "Content-Type": "application/json" },
				body: JSON.stringify(request),
			});

			if (!response.ok) {
				throw new ChatError(
					`HTTP error! status: ${response.status}`,
					"HTTP_ERROR",
					{ status: response.status },
				);
			}

			const reader = response.body?.getReader();
			if (!reader) {
				throw new ChatError("Response body is null", "NULL_RESPONSE");
			}

			return this.createMessageStream(reader);
		} catch (error) {
			if (error instanceof ChatError) throw error;
			throw new ChatError("Failed to fetch chat stream", "FETCH_ERROR", error);
		}
	}

View on GitHub (pinned to 338b89cfe9)

Solutions

1. Open devtools Network tab, take the logged 'Final ChatRequest payload', and replay it with curl to see the server's error body
2. Verify every field of the request (pattern, model, vendor) exists server-side
3. Fix the server-side cause (restore pattern, re-pull model, set API key)

const models = await modelsApi.getAvailable();
if (!models.some(m => m.name === request.model && m.vendor === request.vendor)) {
  throw new Error(`Model ${request.vendor}/${request.model} is not available`);
}

function isChatHttpError(e: unknown): e is ChatError & { code: 'HTTP_ERROR' } {
  return e instanceof ChatError && e.code === 'HTTP_ERROR';
}

try { stream = await chatService.sendMessage(request); }
catch (e) {
  if (isChatHttpError(e)) {
    const status = (e.details as any)?.status;
    if (status === 404 || status === 400) showUserFixableMessage(); // bad pattern/model
    else showServerError();
  } else throw e;
}

When it happens

Trigger: POST /api/chat with a request referencing an unknown pattern/model/vendor (400/404), backend LLM provider unreachable (5xx), or malformed request JSON the chat handler rejects.

Common situations: Pattern renamed or missing server-side; selected model no longer available (Ollama model pulled); API key for the vendor missing server-side; backend upgraded with breaking request-schema changes while the frontend is stale.

Related errors

• HTTP error! status: ${response.status}
• NULL_RESPONSE
• STREAM_CONTENT_ERROR
• errorData.error || `HTTP error! status: ${response.status}`
• await response.text()

AI-assisted analysis of danielmiessler/Fabric@338b89cfe9 (2026-08-15). Data as JSON: /api/errors/b256b90eeea06621. Report an issue: GitHub.