Chatnik brings LLM conversations into Unix shell as native building blocks, letting developers pipe AI chat into existing workflows instead of using separate apps

A Raku package called Chatnik was released that adds CLI (command-line interface) scripts to invoke and maintain conversations with multiple AI language models. Unlike ChatGPT or Claude's web interfaces, Chatnik stores chat history as OS files and integrates directly into Unix shell pipelines, making it composable with grep, awk, sed, and other standard tools.

Instead of locking users into a single app's interaction loop, Chatnik treats LLM calls as Unix primitives—meaning a user can pipe the output of one command (like 'fortune') into an AI prompt, then pipe the result into another tool (like a diagram renderer) in a single shell command. Prompts can be loaded from JSON files and preprocessed using a custom language.

For software engineers and system administrators who already build automation with shell scripts and pipelines, this eliminates the friction of context-switching to separate AI chat apps. A developer can now embed AI-powered copy-editing, code generation, or data transformation directly into existing CI/CD pipelines or daily shell workflows—the same way they already use grep or curl.