Announcements·2 min read
Introducing the Codify Agent
Codify is now agentic. Describe your dev environment in plain English, get a config generated, and apply it - all from a single chat window. No editor required.
Today marks a new chapter for Codify. We're introducing the Codify Agent - and with it, a fundamentally different way to manage your development environment.
Until now, using Codify meant writing a config. You'd open a document, add resources, configure parameters, then apply. That's still there, and it's still great - but it's no longer the only way. With the Codify Agent, you can go from a plain-English description to a fully applied dev environment without ever touching a config file.
Describe what you want. Get a config. Apply it. Done.
A New Way to Use Codify
Open the AI Assistant in the dashboard and just tell it what you need:
- "Set up my machine for Python data science"
- "Set up a Go backend environment with Docker and Kubernetes tools"
- "Update my Terraform to version 1.15.0"
The agent figures out the right tools and versions, asks any clarifying questions it needs (which Python version? personal machine or team setup?), and generates a complete Codify configuration directly in the chat. You can inspect it, tweak it, and when you're ready - apply it, right there, without leaving the window.
No editor. No JSON. Just a conversation.
This is what agentic dev environment management looks like: the agent handles the translation from intent to config, and you stay in control of what actually runs on your machine.
Not Just Another AI Chatbot
You might wonder: why not just ask ChatGPT or another general AI for this? The difference is what the Codify Agent actually knows.
General AI tools generate bash commands from memory. Those commands can be outdated, system-specific, or hard to undo. The Codify Agent works differently - it draws from Codify's full database of validated resources and uses the same plugin system that powers every apply. When a resource exists in the library, the agent uses it. That means pre-tested, declarative, reversible changes - not one-off scripts.
Only for things that aren't covered by the resource library does the agent fall back to a direct script. The result is a setup that behaves exactly like one you'd write by hand, with all the same reliability and safety guarantees.
Also: AI Inside Every Document
For users who do work directly in documents, there's a second agent built into the editor - the Codify Editor Agent. Open any Codify document and you'll find an AI chat panel in the left sidebar, fully aware of your current config.
Ask it things like:
- "Add nvm configured for Node 20"
- "Remove the Terraform block"
- "What does the pyenv resource do?"
Rather than editing the document directly, the editor agent proposes changes as a diff you can review and accept. It's the fastest way to make targeted changes when you're already inside a file.
Get Started
The AI Assistant is available now in the Codify dashboard. If you're an existing user, it's already there in the left navigation. If you're new, you can sign up for free.
We're excited to see what you build with it. As always, feedback and questions are welcome at support@codifycli.com.
Cheers!