Show HN: Boucle – A self-dogfooding autonomous AI agent framework in Rus (github.com)

Boucle, a newly launched framework for developing autonomous AI agents, was unveiled in a project that emphasizes self-dogfooding, as it is built and improved by the very agent that runs on it. This innovative infrastructure allows agents to operate on a persistent schedule, maintaining memory across iterations and making decisions with human-defined boundaries. Boucle moves beyond traditional chatbots by enabling agents to autonomously think, act, learn, and repeat over extended periods, which positions it as a significant advancement in the field of AI/ML. The framework incorporates several key technical features, including a structured loop runner for scheduling, a file-based memory system called Broca, and a Model Context Protocol (MCP) server that facilitates multi-agent collaboration. With built-in goal tracking, comprehensive logging, and robust security measures, Boucle ensures transparency and trustworthiness while allowing for powerful computations without reliance on databases or cloud services. Its open-source nature, combined with a focus on human oversight for critical actions, makes Boucle a promising tool for creating trustworthy autonomous systems that can continuously evolve and improve their capabilities.