This is basically an X-ray for your codebase...
Mindwalk is a new open-source developer tool that turns your codebase and AI coding sessions into an interactive 3D city, making it easier to see how agents such as Claude Code and Codex explore, read, and edit a project.
In this video, we test Mindwalk on a real codebase, replay an AI agent session, follow its path across files, and break down the color-coded timeline that shows where the agent looked, made changes, launched subagents, compacted context, and received new instructions.
🔗 Relevant Links
https://github.com/cosmtrek/mindwalk
❤️ More about us
Radically better observability stack: https://betterstack.com/
Written tutorials: https://betterstack.com/community/
Example projects: https://github.com/BetterStackHQ
📱 Socials
Twitter: / betterstackhq
Instagram: / betterstackhq
TikTok: / betterstack
LinkedIn: / betterstack
📌 Chapters:
00:00 Mindwalk Turns AI Coding Sessions Into a 3D Codebase
00:32 Why AI Agent Logs and Git Diffs Miss the Full Story
01:00 Replaying a Claude Code Session in 3D
01:30 How Mindwalk Tracks Files Read, Viewed, and Edited
02:25 AI Agent Code Review and Scope Analysis
02:49 Mapping Any Git Repository as a 3D City
03:00 Is Mindwalk Actually Useful or Just a Visual Gimmick?
03:32 Current Limitations: Claude Code, Codex, and No Cursor Support
04:06 Who Should Use Mindwalk Today?