Built this because Langfuse does not have native Slack alerting and I kept wanting eval or trace issues to show up where the team already works.
What it does:
- polls Langfuse for new traces and score events
- sends readable Slack alerts
- can also send a generic webhook, so you can trigger n8n workflows when a score drops or a new production trace appears
A few n8n uses I had in mind:
- fan alerts out to email, Discord, or Linear
- open incident workflows when evaluation scores breach a threshold
- route trace events into internal ops automations
It is free to try, and the hosted option is low-cost ($19/mo for unlimited alerts).
Site: https://plugflow.nanocorp.app
If anyone here is doing LLMOps with n8n in the loop, I would love feedback on whether webhook-first is enough or if a dedicated n8n node would be more useful.