Course outline

AcademyModule 0 · Orientation

First launch: your first ten minutes

Lesson 2 of 310 min

What you'll learn

  • You can read the welcome screen and know what each card, button, and thumbnail means
  • You can run the 3-minute guided tour in a sample workspace, with no account and no credits
  • You can open a folder as a workspace and know exactly what to click first
  • You can sign in with Google or an email code, and explain what stays local versus what goes to the cloud

Most apps make you sign up, pick a template, and import everything before you know if they're any good. Consilience goes the other way. The first screen is just a launcher, and the guided tour runs in a sample workspace built on your own disk. You won't need an account until an AI model actually has to read something.

This lesson covers your first ten minutes: the welcome screen, the built-in tour, your first workspace, signing in, and the line between local and cloud.

The real Consilience window, running live. Click in and poke around.

Meet the welcome screen

When the main window opens with nothing to restore, you land on the welcome screen, a launcher. If you've been here before, you'll see a grid of recent workspaces. The picture on each card is a 3D snapshot of that workspace's real knowledge graph, captured ahead of time, and it starts to spin when you hover. Below it sits a line of live counts like "12 entities · 34 notes": entities are the people and concepts pulled from those notes.

  • First-run hero. Titled "New here?", with two buttons: "Take a look around (3 min)" and "Start with my own folder". It shows only on a genuine first run.
  • "Recent workspaces". Up to 5 cards, with pinned ones first. Once you have more than 5, a "View all" link opens the search palette narrowed to workspaces.
  • "+ New workspace". The dashed cell at the end. Its menu offers "Open workspace", "Clone repository", and "Import shared workspace".
  • A gear button labeled "Open settings" in the titlebar. It's the only tool visible on the welcome screen.

Take the 3-minute tour

On a first run, "Take a look around (3 min)" starts the onboarding journey. It opens with a question, "What will you bring in?", and eight role cards: "Research & learning", "Writing & notes", "Code & docs", "Work knowledge", "Trip planning", "Everyday life", "Health & training", "Startup building". Whatever you pick reshapes the whole sample. Consilience writes a real folder of notes into your Documents folder, with a knowledge graph already built. Nowhere in the journey do you need an account or a single AI credit.

  1. "This is a knowledge graph". The sample opens straight onto its graph. Every note behind it is a plain markdown file, and the graph shows how the people, books, and ideas inside them connect.
  2. "Click any node". A spotlight ring drops onto the graph. Each dot is a concept pulled from the documents, and the step finishes only when you open the note behind one.
  3. "Try an example question". Chips badged "Example" replay a full pre-written agent turn: the question, the reasoning, a real search-tool card, and a streaming answer. The point: each answer draws several documents together at once, which keyword search can't do.
  4. "Now it's your turn". The handoff. You get optional capability toggles and two buttons: "Open my own folder" and "Later".

Open your own workspace

Now you point Consilience at a folder of your own. Each window holds one workspace at a time, and the "Home" button, a house icon at the far left of the titlebar, takes you back to the launcher to switch.

Welcome screen
The launcher. Recent-workspace cards plus the "+ New workspace" cell. The Home button brings you back here.
Graph thumbnail
The picture on a workspace card. A snapshot of its real knowledge graph, captured ahead of time, with live entity and note counts.
Empty-workspace command menu
What you land on when a workspace has no open tabs. A task list with live shortcut chips.
Onboarding journey
The 3-minute guided tour inside a generated sample workspace. Offline, the same every time, and free of credits.

Mount your first real folder

  1. Open the picker

    On the welcome screen, click the "+ New workspace" card and choose "Open workspace". The same menu can also "Clone repository" from GitHub or "Import shared workspace" from a .zip bundle.

  2. Pick a folder with something in it

    Choose any folder of notes or documents. If you pick one with no readable documents, it gets added to your recents but you stay on the launcher. That's a guard against dropping you into a blank editor.

  3. Land on the command menu

    The workspace mounts and you see "{your folder} · No open tabs" with a task list. No document opens on its own. Your files wait in the file tree on the left sidebar.

  4. Click "Extract sources"

    While the knowledge graph is still empty, "Extract sources" (⌘⇧I) gets promoted to the one primary button. It starts turning your files into the graph; the extraction lessons later in the course show exactly how.

Quick check

You open a workspace and see a command menu instead of your files. What happened?

Sign in and pick a plan

Everything so far ran without an account. That changes when an AI model has to read your notes, like in chat or extraction, because every AI feature draws on your account's usage allowance. Open Settings with the "Open settings" gear on the welcome titlebar, or hit ⌘, from anywhere, then go to the "Account" tab.

Two options: "Continue with Google", or type your email and click "Email me a code". The 6-digit code verifies itself the instant you finish typing it.

PlanHow the plan picker describes it
Free"Try out AI features at a light pace."
Plus"Plenty for everyday note-taking and agent use."
Pro"Built for heavy, all-day agent work, with early access to advanced features."
Ultra"Top limits for power users, with early access and priority support."

Plans differ only in usage limits. Nothing is locked behind a feature gate. Usage is counted in credits, tokens weighted by what the AI costs, and totted up over two spans: the last 5 hours and the last week.

One thing to know on day one: if you've never paid before, the Plus plan shows "First month free". You do need a card, billing starts when the trial ends, and the offer is once per account.

What runs locally, what runs in the cloud

Consilience is local-first. Your files live in a plain folder you picked, your settings stay on your disk, and the knowledge graph lives on your machine. A workspace can carry its own portable export, .consilience/graph.nq, which rebuilds the graph without re-running the AI. The sample tour is fully offline.

So what leaves your machine? Two things. The AI calls behind chat and extraction run in the cloud, against your account's allowance. And billing happens in your system browser, never inside the app.

Quick check

Which of these works with no account and no AI credits?

The checklist, and one thing to know

After the tour, a "Get started" card settles at the bottom of the Agent sidebar's session list, with a progress bar and five steps. It maps your first day, and the course ahead. Finish all five and it celebrates for a few seconds, then hides for good. "Dismiss" hides it early, and you can bring it back from Help.

  1. "Take the tour" starts the journey you just finished.
  2. "Open a workspace" checks itself off automatically.
  3. "Extract sources" opens ingest. The extraction lessons pick up here.
  4. "View your knowledge graph" opens the graph tab.
  5. "Ask the AI a question" starts a new chat. That's the working-with-the-agent lessons' territory.

While the extraction you just started fills the graph in the background, the next lesson looks at what is being built: the workspace, the graph, and the agent, and how the three fit together.

First launch: your first ten minutes · Consilience Academy