FlowiseAI/Flowise · error · Error

Google requires a tool name for each tool call response, and

Error message

Google requires a tool name for each tool call response, and we could not infer a called tool name for ToolMessage "${message.id}" from your passed messages. Please populate a "name" field on that ToolMessage explicitly.

What it means

Thrown in convertMessageContentToParts (line 323) when a ToolMessage has no name and inferToolNameFromPreviousMessages() cannot match its tool_call_id to a prior AIMessage tool_call. Gemini's API requires each function/tool response to carry the called tool's name, so the converter refuses to emit an unnamed functionResponse.

Source

Thrown at packages/components/nodes/chatmodels/ChatGoogleGenerativeAI/FlowiseChatGoogleGenerativeAI.ts:323

    } else {
        if ('type' in content) {
            throw new Error(`Unknown content type ${content.type}`)
        } else {
            throw new Error(`Unknown content ${JSON.stringify(content)}`)
        }
    }
}

export function convertMessageContentToParts(
    message: BaseMessage,
    isMultimodalModel: boolean,
    previousMessages: BaseMessage[],
    model?: string
): Part[] {
    if (ToolMessage.isInstance(message)) {
        const messageName = message.name ?? inferToolNameFromPreviousMessages(message, previousMessages)
        if (messageName === undefined) {
            throw new Error(
                `Google requires a tool name for each tool call response, and we could not infer a called tool name for ToolMessage "${message.id}" from your passed messages. Please populate a "name" field on that ToolMessage explicitly.`
            )
        }

        const result = Array.isArray(message.content)
            ? (message.content
                  .map((c) => _convertLangChainContentToPart(c as MessageContentComplex, isMultimodalModel))
                  .filter((p) => p !== undefined) as Part[])
            : message.content

        if (message.status === 'error') {
            return [
                {
                    functionResponse: {
                        name: messageName,
                        response: { error: { details: result } }
                    }
                }

View on GitHub (pinned to abe4a8601a)

Solutions

1. Set an explicit name on the ToolMessage matching the tool that was called.
2. Preserve the originating AIMessage (with its tool_calls) earlier in the message array so inference can match tool_call_id.
3. Ensure tool_call_id on the ToolMessage exactly equals the id of the corresponding tool_call in the preceding AIMessage.

Example fix

// before (throws): new ToolMessage({ content: '42', tool_call_id: 'call_1' })  // no name, no prior AI msg
// after:           new ToolMessage({ content: '42', tool_call_id: 'call_1', name: 'calculator' })

import { AIMessage, ToolMessage, BaseMessage } from '@langchain/core/messages'

function toolMessagesNeedName(messages: BaseMessage[]): ToolMessage[] {
  const toolCallsById = new Map<string, string>() // id -> name
  for (const m of messages) {
    if (AIMessage.isInstance(m)) for (const tc of m.tool_calls ?? []) toolCallsById.set(tc.id, tc.name)
  }
  return messages.filter((m): m is ToolMessage =>
    ToolMessage.isInstance(m) && !m.name && !toolCallsById.has(m.tool_call_id)
}
// before invoke:
for (const tm of toolMessagesNeedName(messages)) {
  throw new Error(`ToolMessage ${tm.id} needs a name; set it or restore the prior AIMessage`)
}

function isNamedToolMessage(m: unknown): boolean {
  return ToolMessage.isInstance(m) && typeof (m as ToolMessage).name === 'string' && (m as ToolMessage).name !== ''
}

try {
  return convertMessageContentToParts(message, isMultimodal, previousMessages)
} catch (e) {
  if (e instanceof Error && /Google requires a tool name/.test(e.message)) {
    // set message.name explicitly and retry, or restore the originating AIMessage
  }
  throw e
}

When it happens

Trigger: A ToolMessage with name===undefined is converted, and none of the preceding AIMessages contain a tool_call whose id equals the ToolMessage's tool_call_id.

Common situations: Message history was trimmed/summarized so the originating AIMessage with tool_calls is gone, tool_call_id/name is mismatched between the AI and Tool messages, or a manually constructed ToolMessage omits name.

Related errors

• Unsupported message input
• Invalid media content
• Unsupported source type: ${block.source_type}
• Please provide image as base64 encoded data URL
• Unknown content type ${content.type}

AI-assisted analysis of FlowiseAI/Flowise@abe4a8601a (2026-08-12). Data as JSON: /api/errors/50692c4cd743efd0. Report an issue: GitHub.