Configuration Issues

[nerd8east-a11y]

I was able to stumble my way through the setup/installation (I am at a super-noob level lol) for the project but I am not able to get CAAL to run any of the commands I give it. I have Ollama, Home Assistant and n8n running on separate machines (3 ubuntu servers, one for each project) and as far as I can tell the CAAL machine can connect to all of them (I get green check marks from the setup page after adding tokens and api keys). I know I am missing something, I just do not know what or where to start looking. Sorry, if this is very vague, I am not very knowledgeable about linux/docker buy very good at copying and pasting commands :) Any guidance would be appreciated.

[cmac86]

Hey @nerd8east-a11y what system are you running on? Did you do a CPU install or are you running with a GPU locally? Are you getting any response from CAAL, hearing any speech?

[nerd8east-a11y]

CAAL is running on Ubuntu Server 24.04.3 with a NVIDIA 4070 Super, 32 GB DDR 4 RAM on an AMD Ryzen 9 5950 CPU. I ran the regular install (not the CPU one). After reading more of the docs, I figured out that I needed to publish the n8n workflows and got the proper MCP credentials setup (I think). I can now see CAAL executing the workflows in n8n, so progress was made. I have three workflows exposed to CAAL: calendar_get_events, espn_get_nfl_scores and hass_control, none of them run properly.

• The NFL scores workflow runs very slow and CAAL tends to hallucinate a lot. Sometimes it gets hung up and times out. Sometimes it kicks up multiple executions, one after another. I have to manually stop all of them, or CAAL keeps creating more runs. Today I let CAAL keep trying and it was able to tell me the scores for the Texans game, but it took a while and several failed runs shown in n8n. CAAL is very persistent I will say - lol.

• The calendar events workflows just times out and CAAL just says "I do not have access to your calendar right now". I created a simple n8n workflow to test the credentials and it runs without issues.

• The hass_control one does not execute at all, CAAL just says: "hass control action = turn on, target = office lamp" for example. I have also tested that n8n can access my home assistant instance with the same credentials and it all looks ok.

I know there was an issue with the new version of n8n (related to runners or something like that), Do you know if that may have something to do with my issues? I am running n8n version 2.3.6.

Thanks again.

[cmac86]

What LLM model are you running?

[nerd8east-a11y]

ministral-3:8b

[nerd8east-a11y]

I have Ollama running on a separate machine with an RTX 3090

[cmac86]

Hm that’s odd. I have had good success with ministral-3:8b. Did you customize the prompt at all or using the default? Can you send some logs from the caal-agent container and some chat message examples to see what it’s doing. I’ve never had it try to tools multiple times and when its usually really quick to run a tool especially something like the NFL scores one.

[cmac86]

Ok regarding the runners, you’re right your instance must be set for external runners. So you either need to run a separate n8n runners container or you need to set the environment variable to internal instead of external.

Fix: Add this environment variable to your n8n container:
N8N_RUNNERS_MODE=internal

If you're using docker-compose something like this:

  services:
    n8n:
      image: n8nio/n8n
      environment:
        - N8N_RUNNERS_MODE=internal

If you’re using docker run:

add -e N8N_RUNNERS_MODE=internal to your command.

If CAAL can't see workflows on the local IP, your n8n might only be accessible via the tunnel. Check:

1. Is it running on 5678 and is port 5678 exposed in your docker-compose? (e.g., ports: - "5678:5678")
2. Try adding N8N_HOST=0.0.0.0 to your n8n environment so it listens on all interfaces
3. Check if Ubuntu firewall is blocking: sudo ufw status - if active, run sudo ufw allow 5678

The Cloudflare tunnel setup might have configured n8n to only be reachable externally. Exposing it locally would help with latency.

[nerd8east-a11y]

Ok, I made the changes you suggested and now all workflows are working with very low latency.
Thanks for your help and sorry for taking so much of your time.

[cmac86]

That’s awesome! You’re welcome. Let me know if anything else comes up!

[nerd8east-a11y]

for reference, this is what my yaml files has:

services:
traefik:
image: "traefik"
restart: always
command:
- "--api.insecure=true"
- "--providers.docker=true"
- "--providers.docker.exposedbydefault=false"
- "--entrypoints.web.address=:80"
- "--entrypoints.web.http.redirections.entryPoint.to=websecure"
- "--entrypoints.web.http.redirections.entrypoint.scheme=https"
- "--entrypoints.websecure.address=:443"
- "--certificatesresolvers.mytlschallenge.acme.tlschallenge=true"
- "--certificatesresolvers.mytlschallenge.acme.email=${SSL_EMAIL}"
- "--certificatesresolvers.mytlschallenge.acme.storage=/letsencrypt/acme.json"
ports:
- "80:80"
- "443:443"
volumes:
- traefik_data:/letsencrypt
- /var/run/docker.sock:/var/run/docker.sock:ro

n8n:
image: docker.n8n.io/n8nio/n8n
restart: always
ports:
- "127.0.0.1:5678:5678"
labels:
- traefik.enable=true
- traefik.http.routers.n8n.rule=Host(${SUBDOMAIN}.${DOMAIN_NAME})
- traefik.http.routers.n8n.tls=true
- traefik.http.routers.n8n.entrypoints=web,websecure
- traefik.http.routers.n8n.tls.certresolver=mytlschallenge
- traefik.http.middlewares.n8n.headers.SSLRedirect=true
- traefik.http.middlewares.n8n.headers.STSSeconds=315360000
- traefik.http.middlewares.n8n.headers.browserXSSFilter=true
- traefik.http.middlewares.n8n.headers.contentTypeNosniff=true
- traefik.http.middlewares.n8n.headers.forceSTSHeader=true
- traefik.http.middlewares.n8n.headers.SSLHost=${DOMAIN_NAME}
- traefik.http.middlewares.n8n.headers.STSIncludeSubdomains=true
- traefik.http.middlewares.n8n.headers.STSPreload=true
- traefik.http.routers.n8n.middlewares=n8n@docker
environment:
- N8N_ENFORCE_SETTINGS_FILE_PERMISSIONS=true
- N8N_HOST=0.0.0.0
- N8N_PORT=5678
- N8N_PROTOCOL=https
- N8N_RUNNERS_ENABLED=true
- N8N_RUNNERS_MODE=internal
- NODE_ENV=production
- WEBHOOK_URL=https://${SUBDOMAIN}.${DOMAIN_NAME}/
- GENERIC_TIMEZONE=${GENERIC_TIMEZONE}
- TZ=${GENERIC_TIMEZONE}
- N8N_PYTHON_PATH=/usr/bin/python3
volumes:
- n8n_data:/home/node/.n8n
- ./local-files:/files

volumes:
n8n_data:
traefik_data:

[cmac86]

Thanks for this. And thanks for working through this with persistence. This is a good learning and I’m sure other users will run in to similar issues.