vladignatyev/brain-map-skill

brain-map-skill is a Python tool that takes a folder of Markdown notes and converts them into a single interactive HTML file showing all those notes as a visual graph. The graph draws each note as a dot and connects dots to each other based on links between notes. It works with Obsidian vaults (a popular note-taking app that stores notes as plain text Markdown files) and with exports from a tool called gbrain. The output HTML file is completely self-contained and opens in any browser with no server, no account, and no internet connection required. The repo ships with a prebuilt demo containing 992 fictional notes organized into three themes (work, study, and life) so you can open it immediately and see how it works before touching your own notes. The demo lets you scrub a timeline to watch the note collection grow month by month, filter by theme, search for specific notes, and click any node to see its connections and details. When building a map from your own notes, the script reads the folder structure to determine themes (top-level folders become color-coded groups), reads note tags and subfolders to assign shapes to different note types, and follows wiki-style links between notes to draw connections. Notes with many incoming links appear larger. If your notes have creation date information in their YAML header, the timeline feature works automatically. If not, the script falls back to file timestamps. The tool runs with nothing installed beyond standard Python, though installing optional packages (networkx, numpy, scipy) switches the layout calculation from the browser to Python, which makes large maps with a thousand or more nodes load faster and look cleaner. The script detects which approach to use automatically. It can also be installed as a skill file for AI coding assistants such as Claude Code, Codex, or Cursor, allowing those tools to call the builder directly when asked to generate a knowledge map. The project is MIT licensed.