claude-code-gui
CLAUDE.md1 starsChanged 9 months agoMITView on GitHub
Read @AGENTS.md for architecture patterns, development rules, and agent-specific instructions. Use pnpm (locked at 10.29.1). Never use npm or yarn. Note: AGENTS.md says "Use npm only" — this is incorrect. The project has pnpm-lock.yaml and "packageManager": "pnpm@10.29.1" in package.json. Always use pnpm. Desktop app: Tauri v2 (Rust backend) + React 19 (frontend) + TypeScript Decision: Is data needed across components? → Zustand. Does it persist between sessions? → TanStack Query. All Rust commands are typed via tauri-specta. After adding/changing Rust…
In this repository25 other files it gives its agents.
- AGENTS.mdAGENTS.md
- frontend-designSkill
- rust-async-patternsSkill
- rust-best-practicesSkill
- tauri-v2Skill
- clean-code-inspectorSkill
- unused-code-cleanerSkill
- commitSkill
- complexity-refactorSkill
- component-screenshotSkill
- deep-thinkSkill
- design-checkSkill
- figma-design-pipelineSkill
- figma-to-codeSkill
- issue-startSkill
- issue-updateSkill
- milestone-executeSkill
- milestoneSkill
- milestone-updateSkill
- plannerSkill
- react-refactoringSkill
- specSkill
- story-generatorSkill
- surveySkill
- testSkill
The file
We split this CLAUDE.md into the rules it gives an agent
6 rules. About 0 are common, 1 is distinctive, and 5 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
5
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 5
Read @AGENTS.md for architecture patterns, development rules, and agent-specific instructions.
- Very specific: line 76
All Rust commands are typed via tauri-specta.
- Very specific: line 78
React → Rust: Import from @/lib/tauri-bindings, use commands.* with Result handling
- Very specific: line 80
Never: use string-based invoke() directly
- Very specific: line 85
React Compiler active: No manual useMemo/useCallback needed
- Very specific: line 86
getState() in callbacks: Use useStore.getState() for event handlers, not hook subscriptions
≈ “Common” is an estimate: at least 1 in 5 of the 26,980 CLAUDE.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
CLAUDE.md
This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
Read @AGENTS.md for architecture patterns, development rules, and agent-specific instructions.
Quick Reference
Commands
# Development (ask user to run dev server — don't run it yourself)
pnpm tauri:dev # Run Tauri app with hot reload
pnpm dev # Vite dev server only (no Tauri)
# Build
pnpm build # TypeScript check + Vite build
pnpm tauri:build # Build native application
# Quality (run after significant changes)
pnpm check:all # ALL checks: typecheck, lint, ast-grep, format, rust checks, tests
# Individual checks
pnpm typecheck # TypeScript type check
pnpm lint # ESLint (zero warnings)
pnpm ast:lint # ast-grep architecture rules
pnpm format:check # Prettier format check
pnpm rust:fmt:check # Rust formatting
pnpm rust:clippy # Rust linting
# Testing
pnpm test:run # Run JS tests once
pnpm rust:test # Run Rust tests
pnpm test:all # Both JS + Rust
pnpm test:coverage # Coverage report (60% threshold)
# Auto-fix
pnpm fix:all # Fix lint + format + rust fmt + clippy
# Rust/Tauri
pnpm rust:bindings # Regenerate TypeScript bindings from Rust (tauri-specta)
# Static analysis (periodic)
pnpm knip # Find unused exports/files
pnpm jscpd # Find code duplicationRunning a Single Test
pnpm vitest run src/path/to/file.test.ts # Single test file
pnpm vitest run -t "test name" # By test name
cd src-tauri && cargo test test_name # Single Rust testPackage Manager
Use pnpm (locked at 10.29.1). Never use npm or yarn.
Note: AGENTS.md says "Use npm only" — this is incorrect. The project has pnpm-lock.yaml and
"packageManager": "pnpm@10.29.1"in package.json. Always use pnpm.
Architecture Overview
Desktop app: Tauri v2 (Rust backend) + React 19 (frontend) + TypeScript
State Management Onion
useState (component-local) → Zustand (global UI state) → TanStack Query (persistent/server data)Decision: Is data needed across components? → Zustand. Does it persist between sessions? → TanStack Query.
Rust ↔ React Bridge (tauri-specta)
All Rust commands are typed via tauri-specta. After adding/changing Rust commands, run pnpm rust:bindings to regenerate src/lib/bindings.ts.
- React → Rust: Import from
@/lib/tauri-bindings, usecommands.*with Result handling - Rust → React:
app.emit("event-name", data)→listen("event-name", handler) - Never use string-based
invoke()directly
Critical Performance Rules
- Zustand selectors required:
useUIStore(state => state.value)— never destructure the store - React Compiler active: No manual
useMemo/useCallbackneeded - getState() in callbacks: Use
useStore.getState()for event handlers, not hook subscriptions - These rules are enforced by ast-grep (
pnpm ast:lint)
Code Style
- Prettier: 2 spaces, single quotes, no semicolons, trailing commas (es5), 80 char width
- TypeScript strict mode with
noUncheckedIndexedAccess - Path alias:
@/*→./src/* - i18n: All user-facing strings in
/locales/*.json(en, ko) - Rust:
format!("{variable}")(modern formatting), Tauri v2 APIs only
Key Directories
src/lib/commands/— Centralized command system (actions for menus, shortcuts, command palette)src/store/— Zustand stores (selector pattern enforced)src/hooks/— Custom hooks (enforced by ast-grep to live here)src/components/terminal/— xterm.js 터미널 UI (TerminalPanel, theme)src-tauri/src/commands/— Rust Tauri commandsdocs/developer/— 24+ architecture and pattern guides.ast-grep/rules/— Custom lint rules (no store destructuring, hooks location, no stores in lib)
Local Status
@CLAUDE.local.md
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

