This article will help you understand and troubleshoot errors in Custom Automations that you might see in the Automation reports.
These errors mean the destination URL rejected the request from Sembly. In most cases, you need to adjust the endpoint, authentication, or what your endpoint expects.
Errors you might see:
400 Client Error: Bad Request for url: …
What it means:
Your webhook endpoint is reachable, but it considers the request invalid. Common reasons:
The endpoint expects a different payload shape or headers.
Required fields are missing or have the wrong type.
The URL points to an HTTP trigger (e.g., Power Automate / Logic Apps / n8n / Make / Slack) that is misconfigured or no longer matches the expected schema.
What you can do:
Inspect the webhook handler (Flow, Logic App, Function, n8n workflow, Zap, etc.) and verify:
The HTTP method is POST and the trigger is configured to accept POST.
The request body schema matches what Sembly sends (e.g., JSON structure, required fields)
Temporarily send a test request with a tool like Postman or curl that copies the payload from Sembly (you can log it on your side) to see what the endpoint expects and adjust accordingly.
Check for extra or missing query parameters in the URL; if the URL was regenerated (e.g. new signature or token), update it in Sembly.
Errors you might see:
401 Client Error: Unauthorized for url: …
What it means:
Your endpoint requires authentication (e.g. API key, bearer token, cookie, SharePoint/OneDrive login), and the request from Sembly is not authorized. Either the credentials are missing, invalid, or expired, or the URL you configured points to a page that requires an interactive login.
What you can do:
If your webhook expects an API key or token in a header or query string, make sure:
The key is included in the URL or headers configured in Sembly.
The key has not expired or been revoked.
For endpoints that are actually web UI pages (SharePoint/OneDrive/Docs/Teams/Planner links, etc.):
Replace the URL with a proper API or webhook endpoint that does not require an interactive browser login.
For Microsoft endpoints, use a dedicated HTTP trigger (Logic Apps, Power Automate, Functions) that accepts anonymous or token-based requests, and configure that URL in Sembly.
If your endpoint is behind authentication (e.g., Supabase function, custom API), confirm that:
The auth scheme matches what Sembly sends (e.g. Authorization: Bearer <token>).
Any IP allowlists or firewall rules allow requests from Sembly.
Errors you might see:
403 Client Error: Forbidden for url: …
What it means:
The server understood the request but refuses to authorize it. This often means the caller does not have permission to access the resource, even if authentication is present.
What you can do:
For SharePoint / OneDrive / Teams / document library URLs:
Use a dedicated webhook/HTTP endpoint (e.g., Logic App, Power Automate Flow, or custom API) that has permission to write to the library, and configure that URL in Sembly.
Ensure the identity used by that Flow/App has the correct read/write permissions on the target site or folder.
For other APIs:
Verify the API key or token has the required scopes/permissions to access the resource.
Check any IP allowlists, WAF, or security filters that might block external calls.
If your endpoint enforces additional rules (e.g., tenant or org checks), confirm that requests from Sembly satisfy those rules.
Errors you might see:
404 Client Error: Not Found for url: <https://hooks.zapier.com/hooks/catch/>...
404 Client Error: Not Found for url: https://…n8n…/webhook…
404 Client Error: Not Found for url: https://…/sembly-webhook
What it means:
The webhook URL you configured does not exist (anymore). The endpoint may have been deleted, disabled, or moved, or the URL may have been copied incompletely.
What you can do:
In your webhook provider (Zapier, Make, n8n, custom API, etc.):
Confirm the webhook/HTTP trigger still exists and is active.
Copy the current webhook URL from the provider.
Update your Sembly automation to use the new/valid URL.
If you are using temporary or testing tunnels (e.g., ngrok, trycloudflare) or Railway dev URLs:
Make sure the tunnel or server is running when the webhook is called.
Replace temporary URLs with more stable endpoints for production use.
Errors you might see:
405 Client Error: Method Not Allowed for url: <https://drive.google.com/drive/folders/>...
405 Client Error: Method Not Allowed for url: <https://app.connecteam.com/>
405 Client Error: Method Not Allowed for url: <https://webapp.sembly.ai/collections/personal/>...
What it means:
The URL you configured is not an API/webhook endpoint and does not accept the HTTP method Sembly uses (typically POST). These are often links meant for browsers (file/folder links, Teams/Planner links, app dashboards) rather than machine-to-machine webhooks.
What you can do:
Replace UI or sharing links (Google Drive, OneDrive, Teams, Planner, Sembly collections, etc.) with proper webhook or API endpoints that are designed to receive POST requests.
If you are using a workflow platform (Power Automate, Logic Apps, Make, Zapier, n8n), create an HTTP-triggered workflow and use its URL, not a generic app or folder URL.
Ensure the HTTP trigger is configured to accept POST requests.
Errors you might see:
410 Client Error: Gone for url: <https://hook.eu1.make.com/>...
410 Client Error: Gone for url: <https://cloud.activepieces.com/api/v1/webhooks/>...
What it means:
The webhook endpoint has been permanently removed or disabled by the provider. The URL will not work again.
What you can do:
In your webhook provider (Make, Activepieces, etc.), create a new webhook or HTTP trigger.
Update the webhook URL in Sembly to the new endpoint.
Remove or archive old automations that point to retired URLs.
Error you might see:
413 Client Error: Payload Too Large for url: <https://hooks.airtable.com/workflows/>...
What it means:
The webhook payload from Sembly is larger than what the destination endpoint allows (e.g. Airtable or another service has a maximum request size).
What you can do:
Check whether your automation sends full transcripts or large content and, if possible, reduce the size:
Send only summaries, key items, or metadata instead of full transcripts.
Split long meetings or large payloads into smaller chunks where the destination supports it.
Review the destination’s documentation for maximum payload size and adjust your use accordingly.
If size cannot be reduced, consider an intermediate service that can store large data (e.g. object storage or a database) and only pass references/IDs via the webhook.
These errors indicate problems on the destination server side (your system or third-party service), not in Sembly. The endpoint is reachable but failed while processing the request.
Errors you might see:
500 Server Error: INTERNAL SERVER ERROR for url: …
500 Server Error: Internal Server Error for url: …
What it means:
The destination server encountered an unexpected error while handling the webhook (e.g. exception in your code, unhandled condition, or dependency failure).
What you can do:
Check logs for your service (n8n, Odoo, custom API, etc.) around the time of the failure to see the root cause.
Add validation and error handling in your webhook handler so it returns a 2xx status for valid requests and structured error responses for invalid ones.
If needed, implement retries or queues on your side so transient errors during downstream calls do not cause the whole webhook to fail.
Errors you might see:
502 Server Error: Bad Gateway for url: …
What it means:
An intermediate gateway or proxy in front of your endpoint (load balancer, API gateway, platform edge) failed to forward the request correctly or received an invalid response from your upstream service.
What you can do:
Check your API gateway / reverse proxy / hosting provider (e.g. managed platform or WAF) logs for the specific requests.
Verify that your upstream service is reachable and healthy and that it responds in time.
If you see 502s only during heavy load, consider increasing capacity or adding retry logic on your side.
Errors you might see:
503 Server Error: Service Unavailable for url: …
What it means:
The destination service was temporarily unavailable (maintenance, overload, or outage).
What you can do:
Check the service’s status page (if available) for incidents.
After the service is back, run the affected automations again.
For critical workflows, consider adding retry logic on your side with backoff when your endpoint calls further downstream systems.
These errors mean Sembly could not successfully connect to your webhook endpoint or DNS.
Errors you might see:
HTTPConnectionPool(...): Max retries exceeded with url: ... (Caused by ConnectTimeoutError(... 'timed out. (connect timeout=10)'))
HTTPConnectionPool(...): Read timed out. (read timeout=10)
HTTPSConnectionPool(...): Read timed out. (read timeout=10)
What it means:
Sembly attempted to call your webhook, but the connection or response took too long. Your server may be offline, behind a firewall, slow to respond, or overloaded.
What you can do:
Ensure your server or webhook service is online and reachable from the public internet.
Optimize your webhook handler to respond quickly:
Do heavy processing asynchronously (e.g. push the payload to a queue and return 200 immediately).
Avoid long-running synchronous calls in the webhook request.
Check any firewalls, VPNs, or IP allowlists that might block requests from Sembly.
Errors you might see:
HTTPSConnectionPool(...): Max retries exceeded with url: ... (Caused by NameResolutionError(... Failed to resolve '...' ...))
What it means:
The domain name in your webhook URL cannot be resolved (DNS issue). This usually means the domain is wrong, no longer exists, or is only available in a private network.
What you can do:
Confirm the domain name is spelled correctly and still in use.
Ensure the domain has a public DNS record if you expect Sembly to reach it from the internet.
If the endpoint is only available on a private network, expose a secure public endpoint (e.g. through an API gateway, reverse proxy, or VPN solution) and use that as the webhook URL.
If you’re still having trouble, please contact our support team at [email protected] with details and a screenshot of the error, if possible.
