Show HN: Trigger-tree; see which project docs your AI coding agent reads (github.com)

trigger-tree has been introduced as a groundbreaking tool that tracks which project documentation AI coding assistants read and use. This local, privacy-centric solution offers developers heat maps of documentation usage, a live pulse dashboard, and insights to ensure AI coding agents are guided by relevant information. Traditionally, documentation is created with the hope that AI will utilize it effectively; however, the uncomfortable reality is that many documents go unread, leading to mistakes and inefficiencies. Trigger-tree addresses this issue by providing metrics on document reads, surfacing unused files, and validating if alterations improve performance in guiding AI assistants. The significance of trigger-tree lies in its ability to transform project documentation from passive artifacts into active, monitored infrastructure. By revealing which documents are actually consulted during task execution, teams can make informed decisions about their documentation strategies, enhancing the AI's ability to follow established guidelines correctly. The tool bridges a critical gap in agent observability platforms, which have largely focused on assessing model tokens and traces but overlooked the specific documentation that informs AI behavior. With trigger-tree, developers can optimize their documentation efforts and ensure that their AI models are well-equipped to adhere to their coding conventions and best practices.