MCP server and client connection problem

[radamesbispo]

Bug Description

MCP client do not connect with MCP server .

To Reproduce

1- create a mcp server test with one single tool no Authentication
2- create new workflow with IA agent and MCP client and connect to the MCP server using the webhook SSE generated in the server.
3- when try to make simple call to try, always receive the error:

Issues:
There was a problem loading the parameter options from server: "Could not connect to your MCP server

Expected behavior

work

Operating System

ubuntu

n8n Version

1.88.0

Node.js Version

latest

Database

SQLite (default)

Execution mode

main (default)

[Joffcom]

Hey @radamesbispo,

We have created an internal ticket to look into this which we will be tracking as "GHC-1522"

[netroy]

can you please check if your domain resolves to IPv6 as well?
It's possible that n8n is trying to connect to the IPv6 address, but the server/reverse-proxy isn't configured to listed to IPv6 addresses.

[ZaKull]

Same error, I have the reverse proxy for IPv6 configured correctly.

[netroy]

Can you please try running another MCP client, like the MCP Inspector, and try to connect to the SSE endpoint from that?
This would help us narrow down if the issue is on the server node, or on the client node. 🙏🏽

[salomaoparkour]

I have the same issue, but I'm not using IPv6 at all.

Debug info

core

• n8nVersion: 1.88.0
• platform: docker (self-hosted)
• nodeJsVersion: 20.19.0
• database: postgres
• executionMode: regular
• concurrency: -1
• license: enterprise (production)
• consumerId: 049bf1ed-cfdf-487e-99d2-27dabc4a10e6

storage

• success: all
• error: all
• progress: false
• manual: true
• binaryMode: memory

pruning

• enabled: true
• maxAge: 336 hours
• maxCount: 10000 executions

client

• userAgent: mozilla/5.0 (windows nt 10.0; win64; x64) applewebkit/537.36 (khtml, like gecko) chrome/135.0.0.0 safari/537.36
• isTouchDevice: false

Env variables of my test instance

      - SERVICE_FQDN_N8N_5678
      - 'N8N_EDITOR_BASE_URL=${SERVICE_FQDN_N8N}'
      - 'WEBHOOK_URL=${SERVICE_FQDN_N8N}'
      - 'N8N_HOST=${SERVICE_URL_N8N}'
      - 'GENERIC_TIMEZONE=${GENERIC_TIMEZONE:-Europe/Berlin}'
      - 'TZ=${TZ:-Europe/Berlin}'
      - DB_TYPE=postgresdb
      - 'DB_POSTGRESDB_DATABASE=${POSTGRES_DB:-n8n}'
      - DB_POSTGRESDB_HOST=postgresql
      - DB_POSTGRESDB_PORT=5432
      - DB_POSTGRESDB_USER=$SERVICE_USER_POSTGRES
      - DB_POSTGRESDB_SCHEMA=public
      - DB_POSTGRESDB_PASSWORD=$SERVICE_PASSWORD_POSTGRES

[netroy]

@salomaoparkour can you also please try another MCP client to connect to the SSE endpoint from the trigger node?

[radamesbispo]

i just try the SSE credentials inside the MCP inspector addon in the Visual studio, and the result its in the picture

[radamesbispo]

Can you please try running another MCP client, like the MCP Inspector, and try to connect to the SSE endpoint from that? This would help us narrow down if the issue is on the server node, or on the client node. 🙏🏽

lookslike mcp-server problem