mulmoclaude
Nurture your own AI assistant on your own computer. Local-first and MIT: memories, data and apps stay as plain files in your workspace. Chat summons the right G
What is mulmoclaude?
Nurture your own AI assistant on your own computer. Local-first and MIT: memories, data and apps stay as plain files in your workspace. Chat summons the right GUI — wiki, spreadsheet, chart, form, 3D. Build small apps for an audience of one, no programming required.
Overview of mulmoclaude
What it is
MulmoClaude is an open-source tool for nurturing your own AI assistant on your own computer. It runs locally and stores memories, data, and apps as plain files in your workspace. The assistant is not tied to a single service provider, and a relay server only carries messages in transit. It is released under the MIT license.
Key features
- Local-first operation: memories, data, and apps stay as plain files on your computer.
- Chat summons the right GUI for tasks: markdown, charts, forms, wikis, spreadsheets, or 3D scenes.
- Schema-driven collections, feeds, and plain files for data storage.
- Build small apps without programming knowledge by describing them in everyday language.
- Messaging bridges available for CLI, Telegram, Slack, Discord, LINE, WhatsApp, Matrix, IRC, and Mattermost.
- Supports 8 UI locales (English, Japanese, Simplified Chinese, Korean, Spanish, Brazilian Portuguese, French, German).
Use cases
- Creating and managing a personal wiki that Claude builds and maintains.
- Tracking invoices, restaurant lists, vocabulary practice, or other personal data.
- Generating charts, images, and illustrated guides.
- Scheduling recurring tasks such as daily news digests.
- Ingestion of articles via URL for long-term memory.
- Accessing the same assistant from a phone or messaging app via bridge processes.
Getting started
bashnpx mulmoclaude@latestThe launcher boots the server and opens http://localhost:3001 in your browser. Prerequisites include Node.js 22.22.2+ / 24.15+ / 26+, Claude Code CLI authenticated, and optionally ffmpeg for movie generation, Docker Desktop for sandbox mode, or whisper.cpp for local voice input on macOS. For running from source, see the repository README.
License and pricing
The license is MIT. The project is open source.
Based on the project's public README and repository metadata.