Chloe - Code-first, Private and Reliable AI Agents.

Chloe runs AI agents that chat with you but do their routine work in plain TypeScript, asking a model only where a step needs judgement.

Add a comment

Replies

Best
I built Chloe for developers who want to build AI agents for businesses, especially small businesses. The idea is simple: use code for predictable tasks, and use AI where you need it. For example, an agent handling an invoice might: Use AI to read the invoice. Use code to check the numbers and apply business rules. Ask a person to approve the payment. You shouldn’t need an AI model for every step. Regular code is more predictable and avoids spending tokens on work that doesn’t need AI. Chloe lets you combine both kinds of steps in TypeScript. I also wanted more control over where agents run and what they do. With Chloe, you own the code, tools, skills, and tests that check whether your agent works as expected. You can run it on your own machine or server and decide which outside services it connects to. Running locally doesn’t automatically make everything private. If you use a cloud AI model, the data you send to that model still leaves your machine. Over the past two years, I’ve worked with businesses as a forward deployed engineer—building tools around their actual needs. That experience shaped Chloe: I wanted agents that I could understand, test, and change before handing them over to a business. Chloe is free and open source. It’s my attempt to make that easier. If you build agents for clients or your own business, I’d like to hear from you.