My LLM CLI tool can run tools now, from Python code or plugins

Overview

My LLM CLI tool can run tools now, from Python code or plugins, giving developers a powerful way to connect large language models to real-world workflows directly from the command line. Instead of treating the LLM as a static text generator, you can expose Python functions, scripts, or external services as callable tools that the model can invoke with structured arguments and inspectable results. The CLI lets you define tools in plain Python or load them dynamically via plugins, then orchestrate conversations where the model decides which tools to call and in what order. This enables dynamic automation: fetching data from APIs, querying databases, transforming files, or triggering CI/CD pipelines, all driven by natural language prompts. You retain full control over execution, logging, and safety boundaries while the LLM handles reasoning and orchestration. Designed with hackers and power users in mind, the tool fits into existing terminal-centric workflows. It plays well with environment variables, shell scripts, and version control, making it easy to prototype, debug, and share AI-powered utilities. Whether you’re exploring agent-style interactions, building internal developer tools, or experimenting with function calling standards, this CLI gives you a transparent, debuggable, and extensible foundation for tool-using LLMs.