context-evaluator: Instructions file for Claude Code

CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

---

Development Workflow

After Every Code Modification

MANDATORY: Run tests after each code change to catch regressions early:

bun run test

If any tests fail, you MUST fix them before completing the task. Do not proceed to lint or consider the task done until all tests pass.

At the end of any task, also run lint and fix issues:

bun run lint

Pre-commit Hook

This project uses Husky with a pre-commit hook that runs:

• bun run test - All tests must pass
• bun run lint - Code must pass linting
• bun run typecheck - TypeScript types must be valid

Commits will be blocked if any fails.

---

Common Commands

# Development
bun run dev              # Both API (port 3001) and frontend (port 3000)
bun run dev:api          # API server only
bun run dev:frontend     # Frontend dev server only
bun run dev:cli          # CLI mode with Claude

# Testing & Quality
bun run test             # Run all tests
bun run lint             # Biome check (lint + format)
bun run lint:fix         # Biome auto-fix
bun run format           # Biome format only
bun run typecheck        # TypeScript check

# Production Build
bun run build            # Full production build (frontend + embedded assets + all platform binaries)

# ⚠️ IMPORTANT: Do NOT use build:binaries directly
# build:binaries only compiles binaries without rebuilding frontend
# Always use the full build command above

# Evaluation
bun run evaluate         # Run CLI evaluation shorthand

# Frontend CSS/Style Changes
cd frontend && bun run css:build  # Rebuild Tailwind CSS (required after Tailwind config changes)

Frontend Style Changes

IMPORTANT: When modifying Tailwind configuration (tailwind.config.js) or adding new Tailwind utility classes, you must rebuild the CSS for changes to appear:

cd frontend
bun run css:build    # Rebuild CSS once
# OR
bun run css:watch    # Auto-rebuild on changes