I ran into some Docker issues (networking between containers) and found out the hard way that sysbox does not like Ubuntu 26 but otherwise, was relatively straight forward.
I’ve also opted to use Sonnet-5 instead of Opus and that too has been great.
I’ve been enjoying the n8n mcp with the Pi agent and Ollama which works great, so I was excited to see this AI Assistant, but the docs don’t show it working with Ollama. Is this coming soon?
Hi n8n team, is AWS Bedrock supported as an LLM provider for the AI Assistant? I came across an image that suggests it is supported, but I couldn’t find any documentation about it. If it is supported, could you share which environment variables need to be configured? I wasn’t able to find this information in the docs.
Great to see this coming to self-hosted, even as a pre-release. I run n8n self-hosted on my homelab (Ubuntu, Docker, Cloudflare Tunnel) and it powers real production workflows, WhatsApp agents for clients in the Dominican Republic. I’ll be setting up the Assistant this week following the early instructions and will report back with how it goes on that stack.
This actually looks really useful. I’m mostly curious about the debugging part, because that’s usually where I lose the most time. Definitely going to try this.