visits
required must be zero.
4. **Was it verified rendered?** Drive the actual page (frontend-verify skill /
headless browser) and confirm the new element appears and its controls work
this repository. Two standalone apps plus
shared packages (no workspace/turbo; every package installs independently).
- `frontend/` - Next.js + React product UI on the BoardUI design system.
Conventions live in `frontend/AGENTS.md` and
`frontend/components/foundations/DESIGN-RAMP.md
built SPA is committed in `dist/`, so no Node build
is required. (The Vite/React frontend source and the offline SQL pipeline live in the author's
separate project
GAIA is a proactive personal AI assistant — full-stack Nx monorepo with a Next.js frontend, FastAPI/LangGraph backend, React Native mobile app, Electron desktop app, and Discord/Slack/Telegram bots.
## Engineering Discipline
release does not need an image rebuild; dependency and bundled-example changes do. The frontend selects the ECR tag from the workspace version, and existing Modal sandboxes do not update
Window management helpers
### AI Features
- PR title/description generation using Claude CLI (`src-tauri/src/commands/ai.rs`)
- Frontend wrapper in `src/lib/ai.ts`
- Uses `find_claude_binary()` to locate Claude Code installation
- Gathers git diff
visible run-scoped worker labels may be coarse English role-family names such as `frontend`, `backend`, or `test`, but Meta_Kim does not project those execution labels as durable agents
next sequence is research -> three core documents -> wait for user confirmation -> Spec / tasks -> frontend first with runtime verification -> backend / tests / delivery.
- In SEEAI mode, the next sequence is research -> compact
dependencies (go mod download + pnpm install)
# Development
make dev # Run both backend and frontend in dev mode (parallel)
make backend-dev # Run only backend (go run ./cmd/server)
make frontend
Always read the current function body before replacing.
## Brand & Design
**Read `BRAND_GUIDELINES.md` before any frontend work.** Key rules:
- **One accent color:** Verdigris `#2B6150` (`bg-primary`) — buttons, links, focus rings
- **Background
framework/functions/`: TypeScript SDK for function authors (`Gram.tool()` API, manifest generation, MCP passthrough)
- `client/`: Frontend React application for Gram. Gram Elements — a chat interface that integrates with Gram MCP servers
handoff.js Generate context documents for session handoff between agents
changelog.js In-app changelog data
frontend/
index.html HTML template with {{STYLES}} and {{SCRIPT}} placeholders
styles.css All CSS including dark/light/monokai themes
app.js
A file Claude Code reads at the start of every session. It holds the commands, conventions and warnings the agent needs for this project.
Where does it go?
At the repository root. Claude Code also reads CLAUDE.md files in subdirectories when it works there.
What should it contain?
Build and test commands, the project's layout, conventions that aren't obvious from the code, and mistakes to avoid. Short files tend to work better than long ones.
CLAUDE.md or AGENTS.md?
Claude Code reads CLAUDE.md; most other agents read AGENTS.md. Many projects keep one and point the other at it.