AI tools are powerful, but developer conversations lose context surprisingly fast. After running into this problem repeatedly, I started thinking about a different approach. This is how the idea for contx was born.
Building contx: a Local-First CLI for Project Context Retrieval
Some time ago, artificial intelligence began to significantly change the way we develop software. Like many developers, I started incorporating it into my workflow: reviewing code, refactoring, debugging issues, and exploring solutions faster.
However, I soon ran into a recurring problem.
As conversations with AI became longer, context started to fade. And when I needed to start a new conversation, the problem appeared again: I had to re-explain the project, the architectural decisions, or the structure of the codebase.
In other words, the problem wasn't a lack of information.
The problem was retrieving the right context at the right moment.
That's when a simple idea appeared:
What if there was a tool capable of retrieving and organizing the context of a project so it could be reused when interacting with AI?
At first, I considered building a browser extension. But as the idea evolved, it started to move in a different direction. With the growing adoption of AI APIs and terminal-based developer workflows, a more universal approach seemed more appropriate.
That's how contx was born.
contx is a CLI tool designed to retrieve relevant context from a project quickly and deterministically, following a local-first philosophy and remaining completely independent from whichever AI model the developer chooses to use.
The design of contx follows a few simple principles:
- local-first architecture
- deterministic context retrieval
- AI provider independence
- natural integration with developer workflows
The goal is not to replace AI, but to help developers work with it more effectively.
But... contx is not about AI. contx is about control.