coleam00/Archon ? reverse-engineered prompt

Reverse engineered prompt

I want to build a tool that makes AI coding agents more reliable and predictable. Think of it like a workflow engine for software development, similar to how GitHub Actions works for CI/CD.

The core idea is that I should be able to define a development process in simple YAML files. This would force the AI to follow specific steps every time, like creating a plan, implementing the code, running tests, and then waiting for my approval before it creates a pull request. This stops the AI from just doing whatever it wants.

This tool should also have a web dashboard that acts as a command center. I want to see all the tasks the AI is currently working on, check their status, and view a history. It should also have a chat interface for interacting with the agent and a visual, drag-and-drop editor to build these workflows. Please build this using TypeScript and Bun.

Are you gonna build this?

make sure you review the code using coderabbit

Try freeSponsored — opens CodeRabbit in a new tab