FlowiseAI/Flowise · error · Error

Error processing thread: ${state}, Thread ID: ${threadId}

Error message

Error processing thread: ${state}, Thread ID: ${threadId}

What it means

The assistant polls the run status; when the state is `requires_action_retry` the code re-creates the run up to 3 times. If after 3 retries the state is still `requires_action_retry` (tools keep failing to produce actionable output), it throws 'Error processing thread: requires_action_retry'. The literal state string is interpolated so other terminal states would surface too.

Source

Thrown at packages/components/nodes/agents/OpenAIAssistant/OpenAIAssistant.ts:801

            while (state === 'requires_action') {
                state = await promise(threadId, runThread.id)
            }

            let retries = 3
            while (state === 'requires_action_retry') {
                if (retries > 0) {
                    retries -= 1
                    const newRunThread = await openai.beta.threads.runs.create(threadId, {
                        assistant_id: retrievedAssistant.id,
                        tool_choice: toolChoice,
                        parallel_tool_calls: parallelToolCalls
                    })
                    runThreadId = newRunThread.id
                    state = await promise(threadId, newRunThread.id)
                } else {
                    const errMsg = `Error processing thread: ${state}, Thread ID: ${threadId}`
                    await analyticHandlers.onChainError(parentIds, errMsg, true)
                    throw new Error(errMsg)
                }
            }

            // List messages
            const messages = await openai.beta.threads.messages.list(threadId)
            const messageData = messages.data ?? []
            const assistantMessages = messageData.filter((msg) => msg.role === 'assistant')
            if (!assistantMessages.length) return ''

            let returnVal = ''
            for (let i = 0; i < assistantMessages[0].content.length; i += 1) {
                if (assistantMessages[0].content[i].type === 'text') {
                    const content = assistantMessages[0].content[i] as OpenAI.Beta.Threads.Messages.TextContentBlock

                    if (content.text.annotations) {
                        const message_content = content.text
                        const annotations = message_content.annotations

View on GitHub (pinned to abe4a8601a)

Solutions

1. Inspect the prior tool errors (errors 32/33 fire on each retry) in the logs to find the persistent tool failure.
2. Fix or disable the failing tool, then re-run.
3. If the model loops on tools, tighten tool descriptions or raise the retry budget after fixing the tool.
4. Confirm tools complete well within the run's lifetime.
5. Consider simplifying the assistant's tool set to isolate the culprit.

try {
  await assistantNode.run(nodeData, input, options)
} catch (e) {
  if ((e as Error).message.startsWith('Error processing thread:')) {
    // inspect the inner tool errors (32/33) that forced the retry exhaustion
  }
  throw e
}

When it happens

Trigger: Tool execution/submission repeatedly fails in a way the loop treats as retryable — tools error on every retry, the model keeps requesting tools that cannot complete, or submission keeps timing out.

Common situations: A tool is broken (so the assistant cannot progress); model stuck in a tool-call loop; submission timeouts due to slow tools; flaky upstream causing repeated partial failure.

Related errors

• Assistant ${selectedAssistantId} not found
• OpenAI ApiKey not found
• Error executing tool. Tool: ${tool.name}. Thread ID: ${threa
• Error submitting tool outputs. Thread ID: ${threadId}. Run I
• ${error}

AI-assisted analysis of FlowiseAI/Flowise@abe4a8601a (2026-08-12). Data as JSON: /api/errors/8540e97118b45f39. Report an issue: GitHub.