Frequently Asked Questions

MCP-first means that AI integration isn't an afterthought — it's the architecture. MCP (Model Context Protocol) is an open standard that lets AI tools like Claude query external data sources during a conversation.

Legate Studio runs an MCP server, so any MCP-compatible AI assistant can read and write to your knowledge base as a first-class operation. Your notes aren't just stored — they're directly addressable by AI tools.

The difference from a traditional note app with an "AI feature" is fundamental: in Legate, the AI integration is built into the data model and the server architecture, not bolted on as a chat interface over your notes.

Legate exposes your knowledge base as an MCP server. Configure it in Claude Desktop (or any MCP-compatible client), and your AI assistant gains a set of tools:

• Search your library by keyword or semantic similarity
• Retrieve a specific note by ID or title
• List categories in your knowledge base
• Create new notes from inside the AI conversation
• Browse recent notes

Your personal knowledge becomes part of the AI's working context — not a separate tab you have to switch to, not a copy-paste workflow.

Three things separate Legate from standard note apps:

1. Capture is structured, not raw. When you add a note, AI runs immediately: transcription, categorization, concept extraction, relationship mapping. You don't get a pile of unorganized text — you get structured knowledge from day one.
2. The graph is automatic. Notes connect to each other without manual linking. The knowledge graph builds itself as you capture — you don't maintain it.
3. The whole thing is queryable by AI. MCP integration means your knowledge base is a live memory substrate, not a document archive. AI tools can read from it and write to it during normal use.

When you upload a voice note (called a Motif), Legate runs it through an AI pipeline in sequence:

1. Transcription — speech is converted to text
2. Categorization — which category of your library does this belong to?
3. Concept extraction — what are the key ideas, entities, and themes?
4. Relationship mapping — how does this connect to what you already know?

The result is a structured note in your library — not a raw transcript. The category, connections, and concepts are already there.

Yes. Every note is a node. The relationships between notes form edges. The graph is built automatically as you capture — no manual linking required.

You can explore the graph interactively: force-directed layout, drag nodes around, click to open notes, zoom into clusters. The graph view reveals the shape of what you know — which topics cluster together, which ideas bridge domains, which notes are isolated.

The graph also powers search: semantic search uses embeddings that reflect the same relationships, so finding related notes works conceptually, not just by keyword match.

A Motif is Legate's name for a voice note or text capture — the raw input you feed into your knowledge base. You can record directly in the app, upload an audio file from your phone, or paste in text. Whatever format it arrives in, it goes through the same AI processing pipeline and becomes a structured library entry.

Yes. Your knowledge lives in a SQLite database on a persistent Fly.io volume — isolated per user, not shared with anyone else.

We don't read your notes, train models on them, or use your content for any purpose beyond running the app for you. Notes are private by default — you opt in to sharing specific notes publicly, one at a time.

For full details on what we collect and why, see the Privacy Policy and Security page.

Yes. Your knowledge base is a SQLite file — a standard database format readable by dozens of tools. You can download it directly from your account settings. Your notes are stored as markdown text inside the database, so the content is human-readable with any SQLite browser or export script. No proprietary format, no conversion wizard needed.

Yes. Any note in your library can be published as a public page. Toggle a note to "published" and it gets a clean URL at legate.studio/pub/yourusername/note-slug.

Published notes appear on your public profile at legate.studio/pub/yourusername. Your profile shows your display name, bio, accent color, and the list of notes you've chosen to share. Everything else stays private.

You can unpublish any note at any time — the public URL stops resolving immediately.

Yes. Every public profile gets a per-user RSS feed at /pub/username/feed.xml. There's also a global feed at /feed.xml that aggregates published notes from all users sorted by recency. Both follow the RSS 2.0 standard and work with any feed reader.