Automation platforms such as n8n and Zapier can loop through large lists much faster than a human would. When those workflows call the 101domain MCP Server, they can easily generate hundreds of rapid, concurrent requests to tools like list-domains-tool.
If the combined traffic from your workflow exceeds your API rate limits, the MCP Server will start returning 429 RATE_LIMIT_EXCEEDED (or 429 TOO_MANY_REQUESTS) errors and your workflow may appear to “fail halfway through” the domain list.
Learn more about the 429 RATE_LIMIT_EXCEEDED error, visit our troubleshooting article Why did I receive a 429 Rate Limit Exceeded error from the API?
To keep your workflow stable on large lists, you’ll need to slow down or batch the requests your automation sends to the MCP Server.
Common approaches include:
-
Adding delay / wait nodes between iterations in your loop.
-
Using rate-limit / throttle nodes or built‑in rate controls in n8n, Zapier, or your chosen platform.
-
Adjusting your AI agent’s instructions or prompts so it consolidates work into fewer tool calls instead of making a separate MCP request for every single item.
Different LLMs and model versions may call MCP tools in slightly different ways (for example, some may try to do too many things at once), so you may need to experiment with model choice and prompt wording to find a configuration that works reliably for your use case and stays within your API limits.