Ontology Atlas provides a local-first, open-source solution for understanding a codebase's purpose, structure, and potential impact of changes. It keeps this crucial information as a computable ontology in repository Markdown files, linking product domains and capabilities to implementation evidence and dependencies. This approach ensures that codebase meaning is cloned, branched, reviewed, and shared alongside the code itself, enabling both human developers and AI agents to start from the same shared understanding and collaborate effectively through a visualized map and an MCP server.
主要功能
01Stores codebase ontology directly in repository Markdown files, allowing Git-based versioning and review.
02Provides a desktop application (macOS, Windows beta) and web app for interactive visualization and editing.
03Includes a bundled MCP server for seamless integration with AI coding agents via standard I/O.
04Offers a CLI for scaffolding, validation, impact analysis, and preflight checks.
05Exports compiled graph data to standard formats like JSON-LD and GraphML for external analysis.
0678 GitHub stars
使用案例
01Assessing the impact of proposed code changes before implementation.
02Onboarding new team members by providing an intuitive map of the codebase's architecture and domain.
03Enabling AI agents to query, understand, and update codebase knowledge consistently and accountably.