The product surface
Everything you need to use your data — without surrendering it.
Bring your scattered data into one archive you own. Decide exactly what an AI can see — a whole source or a bounded project. Watch what it does and undo it. Then point any intelligence at it: your own key, the subscription you already pay for, or a model on your own machine.
What an AI actually sees
Each capability is badged for how far along it is — available, preview, or experimental. That's capability maturity, not a shipping promise: Nolocron is a pre-public preview, and anything still on the roadmap is described as roadmap, not as a feature you have today.
Bring your life in
Your knowledge already exists — it's just trapped across two dozen tools, each in its own format. This is where it stops being scattered and becomes one archive you own: read in its real shape, kept on your disk, with a clean way back out.
Two dozen sources, one archive
Datasources
Two dozen kinds
- ChatGPT
- Claude
- Obsidian
- Apple Notes
- iMessage
- Readwise
- + 18 more
Each normalized into one block model — origin kept on every item.
Sources
Pull your chats, notes, files, mail, messages, bookmarks, and highlights out of two dozen apps and into one place you own.
Read explainerImport
A staged pipeline normalizes any source into one model while keeping its provenance and structure intact.
Read explainerInbox
A watch folder that detects, de-duplicates, and imports dropped exports without a single click.
Read explainerCustody
Local-first storage in open formats — SQLite and content-addressed files, with no required account or cloud.
Read explainerExports
Export any slice to open formats — Markdown, CSV, HTML — on demand or on a schedule.
Read explainerFind anything you've kept
Once your sources share one model, the whole archive answers as a single thing. Search across everything at once, read each source in the shape it was meant to have, and start to see how what you know connects.
One search across everything
4 hits · across every source
Search
Full-text search across every source in your archive, with filters, snippets, and optional on-device semantic recall.
Read explainerBrowse
Browse chats as chats, documents as documents, lists as lists — one explorer, many shapes.
Read explainerGraph
An early graph layer over your archive — people, threads, and the relationships between them, traced to evidence.
Read explainerDecide what an AI can see
This is the part nobody else does. Before any model is in the room, you decide what it can reach — scope it to a source or a bounded project, hand-pick a set to hand off, and shut any source off with a switch. It sees only what you grant.
Your sources, your switches
Scope
Grant a whole source or a bounded project — and shut any source off with a switch.
Read explainerCollections
Durable, ordered sets of live block references — curate the slice without copying it.
Read explainerProjects
A project gathers the right sources, scopes an assistant to them, and keeps its own document and threads.
Read explainerSee what it did. Undo it.
Control means nothing if you can't verify it. An assistant works in a sandbox that can never corrupt your real data, every action it takes leaves a receipt, and anything it did can be rolled back. You don't have to trust it — you can check it.
The trail behind one answer
SCOPE
Held to project ‘Q3 Launch’ (locked)
READ
12 blocks from ‘Launch notes’
TOOL
vault.search — query: ‘pricing tiers’
TOOL
block.hydrate — 3 blocks expanded
CITE
4 sources cited in the answer
RESULT
Answer returneddone
RUN RECEIPT
Receipt saved · 7f3a91e2-04bd
Recorded in the append-only event store — reviewable after the fact.
Use any intelligence — or none
Point the brain you already have at your archive — your own API key, the Claude or Codex plan you pay for, or a model running entirely on your machine. The model is rented and interchangeable; the memory it draws on is yours. And you can do the whole job with no AI at all.
Bring any brain — or none
Select model
Assistant
A governed in-app assistant that answers from your knowledge, cites what it used, and leaves a receipt.
Read explainerYour keys
Use the model you trust — OpenAI, Anthropic, Google, or xAI — with your own API keys.
Read explainerSubscriptions
Already subscribed to Claude or Codex? Use it as the engine, with Nolocron as the governed data layer around it.
Read explainerLocal models
On-device models via LM Studio — embeddings and chat that never leave your disk, no API keys.
Read explainerReach it from anywhere
The same governed archive, through whichever door fits the moment — your own AI app over MCP, or a question fired off from Telegram on your phone. One archive, under one set of controls, reachable on your terms.
How the connection works
Your AI app
Claude, Cursor & friends
The chat app you already use, asking in plain language.
The governed gate
The MCP bridge
A small, controlled toolset — not a backdoor to everything.
- search & read
- list & scope projects
- your whole archive
Your vault
Local, on your disk
Your archive in open formats — it never leaves your machine.
On your terms
Bring it in, scope it, watch it, use it.
On your own machine, with a record of everything any AI touched. That's the whole product, and it's yours.