Show HN: Telos – A structured context framework for humans and AI agents (github.com)

A new AI development tool called Telos has been introduced to enhance how both humans and AI agents track the intent behind coding decisions. Unlike traditional systems that primarily track changes in source code through textual diffs, Telos provides a structured framework to capture the reasoning, constraints, and impacts of coding choices. This is crucial as AI agents increasingly take on roles in software development, demanding a system that can convey past decisions and intent—information that is often lost in standard Git commit messages. Telos operates alongside Git, utilizing a content-addressable object model to create a Directed Acyclic Graph (DAG) that mirrors development history. Each design decision, intent, and associated constraints are stored as queryable objects, enhancing cross-session context recovery and facilitating more efficient software development processes. For example, AI assistants can retrieve relevant intents and constraints quickly, which allows them to understand not just what has changed, but why those changes were made. The introduction of Telos represents a significant leap in how development teams can collaborate with AI tools, setting the stage for richer interactions and better-informed code evaluation and debugging processes.