I created a workflow with the API, when i try to open in in the dashboard there is an error: Something Went Wrong! Cannot read properties of undefined (reading 'length') workflow: id:"dfb73a06-11ed-40c8-85a9-5d918f56d481" orgId:"0c4a538b-7cef-44a5-a1a7-b8533e8f45bc"

Hi there! 👋 Sorry to hear you're having trouble opening your workflow in the dashboard. I wasn't able to find the exact cause of the 'length' property error for your specific workflow ID. However, I can help with several troubleshooting steps for this type of error when opening API-created workflows in the dashboard. According to the [troubleshooting information](https://discord.com/channels/1211482211119796234/1211483291191083018/threads/1328923059905630238) I found, here are some steps you can try: ## General Troubleshooting Steps: 1. Refresh your dashboard page or try clearing your browser cache 2. Try incognito mode or a different browser 3. Review browser console and network logs to get more details about the error ## Potential Causes: The error "Cannot read properties of undefined (reading 'length')" typically occurs when the code is trying to access a property of an object that doesn't exist. In your case, this might be due to: 1. **Workflow Validation Issues**: According to [Vapi documentation](https://docs.vapi.ai/workflows), "A workflow cannot be saved until every node is connected and configured." When creating via API, make sure all nodes are properly connected. 2. **Data Type Issues**: The error could be related to [incorrect data types](https://discord.com/channels/1211482211119796234/1211483291191083018/threads/1332319207936757815) being used in the API request that created the workflow. 3. **Analysis Extraction Object Properties**: There might be [inconsistencies in the analysis extraction object properties](https://discord.com/channels/1211482211119796234/1211483291191083018/threads/1339019564838686740) if you created an object with required properties and then removed or renamed them. 4. **Beta Feature Issues**: It's worth noting that [workflows are in beta](https://discord.com/channels/1211482211119796234/1211483291191083018/threads/1341786285379162123) and therefore potentially buggy, especially with conditional nodes.

I was able to figure out what's causing the problem, in the workflow POST request (to create the workflow), I had the following definition (: const workflowDefinition = { name: 'test', nodes: [ { type: 'start', name: 'start', metadata: { position: { x: 120, y: 100 }, }, }, ...] edges: [ ... ] model: { model: 'gemini-2.0-flash', provider: 'google', temperature: 0.7, maxTokens: 4000, }, } If I don't send the model object, the workflow works in the dashboard. For now it's okay for me, but may you could check if this behavior is intended?

Hey, nizmo, you found it right. This is the intended schema or structure for the assistant. Let me know if you require further help.