Skip to content
Agensaë
Mac App Store release candidate macOS Windows

/ 01Desktop app


Sheeter

A local-first cheatsheet viewer for developers — search, copy, return to flow during AI-assisted work. Also the case study for our spec-first agent-native build workflow.

Updated

Sheeter is a lightweight desktop app for browsing, searching, and copying code snippets and reference notes—stored locally on your machine.

Sheeter main workspace with a React cheatsheet loaded and reusable snippets listed in the center pane.
Primary workspace: pick a sheet, scan entries, and copy exactly what you need.
Feature 01 · Stay in flow

Local lookup, not a context switch.

Default hotkey Space

Abstract

When you're deep in an AI-assisted coding session, small interruptions break flow: "What was that flag?" or "What's our project convention for X?" You either waste context tokens on trivia or alt-tab to Google.

Sheeter keeps your cheatsheets one keystroke away — locally, so prompts stay focused on hard problems, not reference material.

What it does

  • Fast search across every registered sheet
  • One-keystroke copy without leaving your workflow
  • Recents stack that surfaces what you used last
  • Section-scoped filters keep results tight
Feature 02 · Library & format

Plain text, your files, your folder.

Sheet formats Markdown + JSON

Abstract

Cheatsheets are Markdown or JSON files you own — version-control them, sync them, diff them, like any other text.

Add Markdown or JSON files through the app, or attach a workspace with a .sheeter/registry.json file.

What it does

  • API reminders, CLI commands, prompt templates, team conventions
  • YAML frontmatter or JSON fields for sections, tags, and display order
  • Files are referenced in place after you add them
  • Re-indexes on launch and on file change
Feature 03 · Local-first by design

No cloud. No telemetry. No account.

Files stay where you choose

Abstract

Sheeter runs entirely on your machine. No account, no analytics, no automatic uploads. Your cheatsheet content stays local unless you export or share it.

Reference your existing Markdown or JSON files in place and search them locally — no upload, no sync server in between.

What this means

  • Your cheatsheet files stay in the locations you choose
  • Teams sync via git or whatever they already use
  • No login, no email required
  • Sensitive snippets stay on your machine unless you choose to export or share diagnostics or content
How it shipped · Agent-native workflow

A workflow tuned for AI coding agents.

Source of truth AGENTS.md

Approach

Agent-assisted development can move fast, but only if the project structure is unambiguous — clear conventions, concrete acceptance criteria, and tight boundaries on every change.

Sheeter is the case study where we tuned that workflow into something repeatable.

What we changed

  • An AGENTS.md contract describing precedence rules and constraints
  • Issue prompts written as self-contained implementation briefs
  • Spec-first docs (technical requirements, blueprint, milestone plan) as the source of truth
  • Cross-platform Tauri foundations chosen for reviewable, scoped change surface

Who it’s for

  • Developers who routinely need quick syntax and command lookups
  • Teams who want a shared, searchable set of engineering conventions
  • Anyone pairing with an AI assistant and trying to avoid burning context on “trivia”

More views

Sheeter search input active while filtering snippets and commands.
1 of 3

Live search filters commands and snippets as you type.

Technical specs

The honest fine print.

Everything a developer would ask before installing a thing on their primary machine.

Platforms
macOS Apple Silicon · Windows
Runtime
Tauri (Rust) + React + TypeScript
Cheatsheet format
Markdown with YAML frontmatter, or JSON
Storage
OS app data for settings/cache; cheatsheet files remain in user-selected locations.
Hotkey
Configurable · default ⌥Space
Telemetry
None — no analytics, no remote logging
Account required
No
Status
Mac App Store release candidate

/ 04 — Next step

Have a similar project in mind?

Tell us your context — we'll help you scope the next step.