Open · local-first · built for real work

A workspace built for people and agents.

Turn files and business data into documents, workflows, dashboards, and applications through visible, reviewable changes—while keeping the underlying work portable.

View the open-source project →

Local-first Open formats Reviewable agent work Complete export
The Lattice unit cell carrying a signal The seven-node Lattice mark glows inside an axonometric grid of addressable resources. real directory semantic command open resource

Available in the local app

Markdown pages & quick captureLinks, backlinks & hybrid searchEditable JSON CanvasSQLite data apps & saved viewsDuckDB, Parquet, charts & mapsNative and Pyodide notebooksTasks, workflows & proposalsCLI, local API & MCP foundations

One working environment

Your work is scattered. Your AI context should not be.

Files live in folders and cloud drives. Operational data lives in spreadsheets and databases. AI work disappears into private chats. Lattice gives people and agents one place to work across the information you already have.

Before

Search, copy, paste, re-explain.

Every new task begins by reconstructing context from disconnected tools.

In Lattice

Connect, act, review.

Work against real resources and inspect what an agent used, did, and changed.

After

Keep durable work.

The output remains a document, table, workflow, dashboard, or application you own.

Agents that work visibly

Review the work, not just the answer.

Lattice gives agents a working environment across files, tables, datasets, and tools. They can gather context, run bounded work, and propose durable changes without silently rewriting the workspace underneath you.

  • See the resources and versions used as context.
  • Follow tool execution, evidence, and generated outputs.
  • Accept, reject, or refine proposed changes before they become canonical.

See agents and automation →

Desktop GUI CLI Local API MCP Agents & scripts Command + Transaction *.md *.sqlite *.parquet *.canvas

People, agents, and tools meet at the same governed command core

Connected, not flattened

Every resource keeps its strengths.

Pages remain great documents. Tables use a real database. Analytical data stays columnar. Notebooks remain Jupyter-compatible. Links, embeds, search, history, and one command system make them feel like parts of the same place.

  • Write and think with the speed of a local-first knowledge tool.
  • Build structured systems without moving the work into a proprietary database.
  • Give agents a versioned environment instead of another disposable chat.

Tour every workspace surface →

Lattice resource constellation Pages, canvases, tables, datasets, notebooks, and tasks joined in one local workspace. Page Canvas Table Dataset Notebook Task Workspace

A real directory

The app is optional. Your work is not.

A Lattice workspace is an ordinary folder. Open it in Finder, Git, VS Code, a backup tool, or another format-aware app. There is no export ceremony between you and your own information.

  • Markdown for pages, JSON Canvas for spatial work, and ipynb for notebooks.
  • SQLite for mutable tables; Parquet and DuckDB for analytical data.
  • External edits are detected and reconciled instead of silently overwritten.

Understand the workspace model →

Engineering Workspace/ lattice.yaml manifest README.md markdown Notes/ pages CRM.data/ data app · sqlite Usage.dataset/ parquet · duckdb Strategy.canvas json canvas Analysis.ipynb jupyter Architecture.ink/ ink · arrow .lattice/ derived

One workspace, visible on disk

Local first. Cloud when you want it.

Start on your computer. Add the connected layer when work becomes shared.

The local app is useful without an account or network connection. Lattice Cloud is in development for managed storage, synchronization, sharing, collaboration, hosted agents, automation, and publishing.

Local workspace: available now
Managed cloud: in development

Explore Lattice Cloud →

Local workspace with optional cloud sync A local file workspace remains canonical while an optional managed cloud connects devices and collaborators. Local workspace · always yours Lattice Cloud · in development sync · sharing · backups · teams

Build on the work you already have

Start locally. Bring in a team when you’re ready.

Join the macOS preview, follow the public project, or apply to shape the first Lattice Cloud team workflows directly with the founder.