n8n-io/n8n · error · Error

Failed to convert to Langchain content block

Error message

Failed to convert to Langchain content block: ${JSON.stringify(block)}

What it means

toLcContent() maps an n8n content block to a Langchain content block by checking block types in turn (text, reasoning, file, tool_call, invalid_tool_call, tool_result, citation, provider/non_standard). If none match, the block is of an unknown/unsupported type and cannot be represented in the Langchain content model, so it throws with the serialized block for diagnosis.

Solutions

Inspect the serialized block in the error to see its shape, then construct it using a supported type guard (text/reasoning/file/tool_*).
If it's a genuinely new block type, add a branch + isN8nXBlock predicate to toLcContent in message.ts.
Sanitize upstream: drop or coerce unknown blocks before conversion.

Example fix

// before
{ type: 'mystery', payload: {...} }

// after (coerce to text or drop)
{ type: 'text', text: JSON.stringify(rawBlock) }

Defensive patterns

Strategy: type-guard

Validate before calling

import { isN8nTextBlock, isN8nReasoningBlock, isN8nFileBlock } from '@n8n/ai-utilities';
function isKnownBlock(b: unknown): boolean {
  return isN8nTextBlock(b) || isN8nReasoningBlock(b) || isN8nFileBlock(b) /* or other supported guards */;
}
content.filter(isKnownBlock).map(toLcContent);

Type guard

function isSupportedBlock(b: unknown): boolean {
  return (
    typeof b === 'object' && b !== null && 'type' in b &&
    ['text','reasoning','file','tool_call','invalid_tool_call','tool_result','citation','non_standard'].includes((b as { type: string }).type)
  );
}

Prevention

Filter content through known type guards before converting.
When adding a new n8n block type, add a branch + predicate to toLcContent the same commit.
Log unknown block types so converter gaps surface in dev.

When it happens

Trigger: Passing a content block missing a recognized discriminator (e.g. `{ type: 'something_new' }`); a malformed block where the type guard predicates all return false; a newly added n8n block type that this converter has not been taught to handle.

Common situations: A provider returned a block shape the converter doesn't know yet (version lag); a tool/handler emitted an ad-hoc block object; schema drift between n8n message types and the converter.

Related errors

AI-assisted analysis of n8n-io/n8n@5ac6606e81 (2026-08-12). Data as JSON: /api/errors/86c6d236a49d4589. Report an issue: GitHub.

Appendix: source

Thrown at packages/@n8n/ai-utilities/src/converters/message.ts:317

	}
	if (isN8nCitationBlock(block)) {
		return {
			type: 'citation',
			source: block.source,
			url: block.url,
			title: block.title,
			startIndex: block.startIndex,
			endIndex: block.endIndex,
			citedText: block.text,
		} as unknown as LangchainMessages.ContentBlock;
	}
	if (isN8nProviderBlock(block)) {
		return {
			type: 'non_standard',
			value: block.value,
		} as LangchainMessages.ContentBlock.NonStandard;
	}
	throw new Error(`Failed to convert to Langchain content block: ${JSON.stringify(block)}`);
}

export function toLcMessage(message: Message): LangchainMessages.BaseMessage {
	const lcContent = message.content.map(toLcContent);

	switch (message.role) {
		case 'system':
			return new LangchainMessages.SystemMessage({
				content: lcContent,
				id: message.id,
				name: message.name,
			});
		case 'user':
			return new LangchainMessages.HumanMessage({
				content: lcContent,
				id: message.id,
				name: message.name,
			});

View on GitHub (pinned to 5ac6606e81)