This gives Claude persistent memory across conversations using a local SQLite database. It's straightforward: search for what you've stored, add new memories with tags, delete what you don't need. The full-text search means you can actually find things later, and you can attach images to memories. The real value is when you're working on a project over multiple sessions and don't want to keep re-explaining preferences, architectural decisions, or deployment commands. It pipes well with other tools and outputs JSON when you need it. Think of it as a notebook that Claude can read and write to, stored in `~/.mem/mem.db` on your machine.
npx -y skills add runablehq/memory --skill mem --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