Claude gets persistent memory across sessions with this one. Everything you discover, decide, or fix gets stored in a single `.claude/mind.mv2` file that you can search semantically, commit to git, or share with teammates. It auto-categorizes memories into types like decisions, solutions, and patterns, then lets you query them with natural language. The selling point is portability: your entire context history lives in one file instead of scattered across chat logs or documentation. If you've ever wished Claude remembered what you figured out last week or wanted to onboard someone by literally handing them your AI's brain, this does that.
npx -y skills add memvid/claude-brain --skill memory --agent claude-codeInstalls into .claude/skills of the current project.
Select a file.
juliusbrussee/caveman
mattpocock/skills
shadcn/improve
obra/superpowers
forrestchang/andrej-karpathy-skills
vercel-labs/skills