Option for customize MCP Connection Time Out

Feature Requests
,
1

The idea is:

Option to set time out for MCP Client Tool or completely remove the 60s time out.

My use case:

I have a Call sub workflow tool node, it call another workflow to generate a full blog content and post it on my blog. The sub workflow takes around 8m to finish the job. Before, when I connected it directly to AI Agent node as a tool, the AI Agent always waited for the finishing of the sub workflow to receive the result. Now I connect the sub workflow through an MCP connection, the MCP client tool node is always timed out after 60s and showing the 32001 error.

I think it would be beneficial to add this because:

This is good for sub-workflows with processing time too long.

Any resources to support this?

No

Are you willing to work on this?

I have no knowledge in coding

2

I have the same problem

3

Please!! This makes MCP tools which have processing time really difficult to use. It will repeatedly call them causing at least double resources and API requests being used.

4

Hi all, does anybody knows about a workaround to increase MCP client timeout running n8n docker?

2025-10-10T13:00:09.229Z | debug | Running node “MCP Client” started {“node”:“MCP Client”,“workflowId”:“MPvockKRcVmtN4Qb”,“file”:“logger-proxy.js”,“function”:“exports.debug”}
2025-10-10T13:00:09.306Z | debug | Original PATH: /usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin {“file”:“McpClient.node.js”,“function”:“execute”}
2025-10-10T13:00:09.307Z | debug | Transport created for MCP client command: npx, PATH: /usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin {“file”:“McpClient.node.js”,“function”:“execute”}
2025-10-10T13:01:00.008Z | debug | Querying database for waiting executions {“scopes”:[“waiting-executions”],“file”:“wait-tracker.js”,“function”:“getWaitingExecutions”}
2025-10-10T13:01:09.540Z | error | MCP client connection error: MCP error -32001: Request timed out {“file”:“McpClient.node.js”,“function”:“execute”}

Many thnk in advance

Viloga

5
6

We now have the opp to adapt the timeout.
BUT: there is no error handling! It would be VERY helpful not only to raise an error but also to sent this error back to Agent node and its errorhandling.