Skip to content
supastack
Browse documentation
DesktopTauri

For AI agents: fetch the complete documentation index at /llms.txt, the full documentation corpus at /llms-full.txt, the agent skill at /skill.md, or this page as Markdown at /docs/desktop/agent-setup.md (or send Accept: text/markdown).

Let an agent set it up

Paste one ready-made prompt into a coding agent, let it run the Desktop setup on your computer, and check the report it writes about each step.

Ask about this guide
Open in

You can delegate the boring part

This repository was written so a coding agent can read it. It ships a map file (llms.txt), a rules file (AGENTS.md), per-file memory notes (.ai/) and task skills (.agents/skills/). Open a terminal inside your copy of the repository to run it — the standalone Desktop checkout does not include the prompts/ folder, so the ready-made prompt is printed in this guide instead of a file.

Delegating the typing does not mean skipping the checks. The agent works inside your copy of the repository and must report what it did with evidence. The sections below show what a good report looks like, so you stay in charge of the result.

This path sets up the Tauri shell and its connection to the Web backend — the server that handles accounts and data. The desktop window only displays that backend; if your checkout also contains apps/web, the agent can run it locally, and a Desktop-only checkout needs the reachable backend address you were given.

What you need first

Finish the preparation guide first: Node.js, Git and an editor installed, the GitHub invitation accepted, and the supastack folder downloaded to your computer. The agent works inside that folder, so everything in Before you start still applies.

You also need a tool that can act as a coding agent — an editor with a built-in AI coding agent (an AI IDE), or an agent that runs in your terminal (an agent CLI). Any current agent-capable tool works; the prompt already spells out what it must read and run, so you do not need to write instructions yourself.

Finally, decide two things the prompt asks for: the backend address the window should load — http://localhost:3000 for a local backend or your provided HTTPS address — and whether Rust and the native build tools are already installed on this computer.

Copy the prompt, fill in two blanks, paste

The prompt is the code block below — select and copy the whole block.

Two placeholders need your choices. Replace <APP_URL> with the backend address the window should load — http://localhost:3000 for a local backend or your provided HTTPS address. Replace <TOOLCHAIN: present | install> with present when rustc and cargo already print versions, or install when the agent should set up Rust and the OS build tools first. The numbered steps stay exactly as written.

Set up this supastack Desktop checkout end to end and open the Tauri window against the backend at **<APP_URL>**.
1. Read llms.txt, AGENTS.md and docs/agent-security.md before changing anything.
2. Run rustc --version and cargo --version. On **<TOOLCHAIN: present | install>** = install, install Rust and this OS's Tauri prerequisites first, then continue.
3. Run corepack pnpm install.
4. If apps/web exists in this checkout, start it in a second terminal; otherwise verify <APP_URL> answers from this computer and report its status.
5. Start the shell with corepack pnpm --filter @supastack/desktop dev and report whether the window opens, the tray icon appears and the summon shortcut works.
6. Stop after any step fails twice and report the failing command and first error. Ask before deleting anything.

What the agent does with the prompt

The prompt points the agent at the map, the rules and the security checklist, then tells it to work through the setup in order and prove each phase. In plain terms, the agent:

  • Reads the repository map (llms.txt), the project rules (AGENTS.md) and the credential-handling rules (docs/agent-security.md) before touching anything.
  • Checks the Rust toolchain and this OS's Tauri prerequisites, and reports what it installs rather than failing halfway through a compile.
  • Verifies the backend answers at your address — starting apps/web itself only when that folder exists in this checkout — instead of assuming a server that is not there.
  • Installs dependencies, compiles the shell and opens the window, then reports window, tray and summon-shortcut behavior with the actual output.
  • Stops and reports after the same step fails twice — it does not work around a failure or delete anything without asking.

What you check in the agent's report

A good report shows evidence, not claims. Look for the backend URL answering, the dependency install completing, a finished Rust compile and an open window behind each PASS — not a sentence like 'everything works'.

The prompt tells the agent to stop and report after the same step fails twice — not to work around the failure — and to ask you before anything destructive. If your agent keeps improvising after repeated failures, stop it and take the first error to the troubleshooting guide.

Two things must never appear: a security feature switched off to silence an error, and a secret value such as a password or key printed into the chat or the report. If you see either, treat the run as failed.

Keep the shell and real sign-in separate in your head. The window showing your backend's homepage proves the shell — real accounts still need the backend's database. 'The window opened' is not yet 'accounts work'.

If you would rather do it by hand

Every step the agent performs is written out in the manual guides below — the same commands, the same checks, at your pace.