English

Neuromapa (English summary)

Claude Code's memory as a live 3D brain. Neuromapa gathers the notes Claude Code keeps about your projects (~/.claude/projects/*/memory/, your CLAUDE.md files and any documents you choose) and draws them as a brain: each note is a neuron, and each time a note mentions another one, a connection. While Claude works, you see what it reads and edits, in which chat, and which agents it launched.

It also:

• checks your notes (/neuromapa:salud): links to notes that no longer exist, repeated memories, and, if you add your projects, file:line references that moved and commits that don't exist;
• finds what you already know (/neuromapa:sobre topic): which notes to read first, outdated facts, recorded decisions and the code those notes cite, so Claude checks it before answering;
• learns from your use: files and notes used together get connected, and what isn't used fades away;
• helps keep the memory tidy: it can split a note that got too long, list what your notes say is still pending, and tell you what a note said before and when it changed. It shows the plan first and writes only with your OK.

Everything runs on your computer. There is no telemetry and no server of its own.

Install

You need Claude Code, Python 3.10 or newer and git. Then:

claude plugin marketplace add lumiel777/neuromapa
claude plugin install neuromapa@neuromapa

Close and reopen Claude Code, then try /neuromapa:abrir (the map), /neuromapa:salud (the check-up) and /neuromapa:sobre <topic> (the search). /neuromapa:abrir --demo shows the map with made-up notes.

More

• The full README in English.
• The rest of this wiki is in Spanish: Inicio.
• Problems: open an issue. Security problems go through SECURITY.md, not a public issue.

Neuromapa is in testing (version 0.1) and is open source under the MIT license.