Codela - AI coding agents that learn from production mistakes
by•
When you deploy, Codela checks your connected monitoring tool (Datadog, Grafana etc) to see if that deploy made your app slower or added errors. If it did, Codela writes a lesson into your repo that Claude Code, Cursor and Codex read before touching that code again.

Replies
Hi Product Hunt! I’m Zach.
I built Codela because AI coding agents keep repeating mistakes production already taught the team. A change ships, latency doubles, someone fixes it, and weeks later an agent writes the same thing again.
So Codela writes those lessons automatically from your deploys and Git history, into .codela/lessons.md, read via CLAUDE.md and AGENTS.md so any agent (claude, codex and cursor) always reads it.
Try it without signing up: paste any public repo at usecodela.com. I’d love your honest feedback, especially on lessons that are wrong or obvious.