Files
golfgame/CLAUDE.md
adlee-was-taken 121959ba27
All checks were successful
Build & Deploy Staging / build-and-deploy (release) Successful in 29s
docs(CLAUDE.md): staging deploy verification checklist
Encodes the lessons from the v3.3.5 → v3.3.5.1 hotfix cascade: CI green
is necessary but not sufficient. Walk the chain: clean worktree → tag
sha matches → container up recently → new code introspectable → env vars
present → DB state correct → end-to-end smoke.

Each step calls out a specific failure mode we just hit, so future-me
doesn't assume the next deploy will 'just work' when the primitives
underneath (git fetch tag-cache, compose env wiring, image reuse) can
silently skip changes.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-04-18 14:06:37 -04:00

337 lines
14 KiB
Markdown

# Golf Card Game - Project Context
A real-time multiplayer 6-card Golf card game with CPU opponents and smooth anime.js animations.
## Quick Start
```bash
# Install dependencies
pip install -r server/requirements.txt
# Run the server
python server/main.py
# Visit http://localhost:8000
```
For full installation (Docker, PostgreSQL, Redis, production), see [INSTALL.md](INSTALL.md).
## Architecture
```
golfgame/
├── server/ # Python FastAPI backend
│ ├── main.py # HTTP routes, WebSocket server, lifespan
│ ├── game.py # Core game logic, state machine
│ ├── ai.py # CPU opponent AI with timing/personality
│ ├── handlers.py # WebSocket message handlers
│ ├── room.py # Room/lobby management
│ ├── config.py # Environment configuration (pydantic)
│ ├── constants.py # Card values, game constants
│ ├── auth.py # Authentication (JWT, passwords)
│ ├── logging_config.py # Structured logging setup
│ ├── simulate.py # AI simulation runner with stats
│ ├── game_analyzer.py # Query tools for game analysis
│ ├── score_analysis.py # Score distribution analysis
│ ├── routers/ # FastAPI route modules
│ │ ├── auth.py # Login, signup, verify endpoints
│ │ ├── admin.py # Admin management endpoints
│ │ ├── stats.py # Statistics & leaderboard endpoints
│ │ ├── replay.py # Game replay endpoints
│ │ └── health.py # Health check endpoints
│ ├── services/ # Business logic layer
│ │ ├── auth_service.py # User authentication
│ │ ├── admin_service.py # Admin tools
│ │ ├── stats_service.py # Player statistics & leaderboards
│ │ ├── replay_service.py # Game replay functionality
│ │ ├── game_logger.py # PostgreSQL game move logging
│ │ ├── spectator.py # Spectator mode
│ │ ├── email_service.py # Email notifications (Resend)
│ │ ├── recovery_service.py # Account recovery
│ │ └── ratelimit.py # Rate limiting
│ ├── stores/ # Data persistence layer
│ │ ├── event_store.py # PostgreSQL event sourcing
│ │ ├── user_store.py # User persistence
│ │ ├── state_cache.py # Redis state caching
│ │ └── pubsub.py # Pub/sub messaging
│ ├── models/ # Data models
│ │ ├── events.py # Event types for event sourcing
│ │ ├── game_state.py # Game state representation
│ │ └── user.py # User data model
│ └── middleware/ # Request middleware
│ ├── security.py # CORS, CSP, security headers
│ ├── request_id.py # Request ID tracking
│ └── ratelimit.py # Rate limiting middleware
├── client/ # Vanilla JS frontend
│ ├── index.html # Main game page
│ ├── app.js # Main game controller
│ ├── card-animations.js # Unified anime.js animation system
│ ├── card-manager.js # DOM management for cards
│ ├── animation-queue.js # Animation sequencing
│ ├── timing-config.js # Centralized timing configuration
│ ├── state-differ.js # Diff game state for animations
│ ├── style.css # Styles (NO card transitions)
│ ├── admin.html # Admin panel
│ ├── admin.js # Admin panel interface
│ ├── admin.css # Admin panel styles
│ ├── replay.js # Game replay viewer
│ ├── leaderboard.js # Leaderboard display
│ └── ANIMATIONS.md # Animation system documentation
├── docs/
│ ├── ANIMATION-FLOWS.md # Animation flow diagrams
│ ├── v2/ # V2 architecture docs (event sourcing, auth, etc.)
│ └── v3/ # V3 feature & refactoring docs
├── scripts/ # Helper scripts
│ ├── install.sh # Interactive installer
│ ├── dev-server.sh # Development server launcher
│ └── docker-build.sh # Docker image builder
└── tests/e2e/ # End-to-end tests (Playwright)
```
## Key Technical Decisions
### Animation System
**When to use anime.js vs CSS:**
- **Anime.js (CardAnimations)**: Card movements, flips, swaps, draws - anything involving card elements
- **CSS keyframes/transitions**: Simple UI feedback (button hover, badge entrance, status message fades) - non-card elements
**General rule:** If it moves a card, use anime.js. If it's UI chrome, CSS is fine.
- See `client/ANIMATIONS.md` for full documentation
- See `docs/ANIMATION-FLOWS.md` for flow diagrams
- `CardAnimations` class in `card-animations.js` handles everything
- Timing configured in `timing-config.js`
### State Management
- Server is source of truth
- Client receives full game state on each update
- `state-differ.js` computes diffs to trigger appropriate animations
### Animation Race Condition Flags
Several flags in `app.js` prevent `renderGame()` from updating the discard pile during animations:
| Flag | Purpose |
|------|---------|
| `isDrawAnimating` | Local or opponent draw animation in progress |
| `localDiscardAnimating` | Local player discarding drawn card |
| `opponentDiscardAnimating` | Opponent discarding without swap |
| `opponentSwapAnimation` | Opponent swap animation in progress |
| `dealAnimationInProgress` | Deal animation running (suppresses flip prompts) |
**Critical:** These flags must be cleared in ALL code paths (success, error, fallback). Failure to clear causes UI to freeze.
**Clear flags when:**
- Animation completes (callback)
- New animation starts (clear stale flags)
- `your_turn` message received (safety clear)
- Error/fallback paths
### CPU Players
- AI logic in `server/ai.py`
- Configurable timing delays for natural feel
- Multiple personality types affect decision-making (pair hunters, aggressive, conservative, etc.)
**AI Decision Safety Checks:**
- Never swap high cards (8+) into unknown positions (expected value ~4.5)
- Unpredictability has value threshold (7) to prevent obviously bad random plays
- Comeback bonus only applies to cards < 8
- Denial logic skips hidden positions for 8+ cards
**Testing AI with simulations:**
```bash
# Run 500 games and check dumb move rate
python server/simulate.py 500
# Detailed single game output
python server/simulate.py 1 --detailed
# Compare rule presets
python server/simulate.py 100 --compare
```
### Server Architecture
- **Routers** (`server/routers/`): FastAPI route modules for auth, admin, stats, replay, health
- **Services** (`server/services/`): Business logic layer (auth, admin, stats, replay, email, rate limiting)
- **Stores** (`server/stores/`): Data persistence (PostgreSQL event store, user store, Redis state cache, pub/sub)
- **Models** (`server/models/`): Data models (events, game state, user)
- **Middleware** (`server/middleware/`): Security headers, request ID tracking, rate limiting
- **Handlers** (`server/handlers.py`): WebSocket message dispatch (extracted from main.py)
## Staging Deploy Verification Checklist
After any release that triggers a staging deploy (via `.gitea/workflows/deploy-staging.yml`), do NOT trust "CI went green" — walk the full chain end-to-end. A green CI run does not prove the deploy did what you intended: `git fetch` won't update already-cached tags, compose yaml and `.env` can drift out of sync, and container images can cache without visible signal. The v3.3.5 → v3.3.5.1 saga cost us two releases because each of these bit in turn.
**Run through every step before declaring a staging deploy successful:**
1. **Worktree is clean on staging BEFORE cutting the release.**
```bash
ssh root@staging.golfcards.club 'cd /opt/golfgame && git status --short'
```
Must be empty. Dirty files or untracked files will abort `git checkout $TAG` mid-pipeline. If you ever scp files to staging for hot-patching, land those changes on main + commit before the release, OR `git reset --hard HEAD && git clean -fd` on staging first.
2. **Staging is actually at the new tag, not a stale cached position.**
```bash
ssh root@staging.golfcards.club 'cd /opt/golfgame && git rev-parse HEAD && git log --oneline -1'
```
Compare the sha to `git rev-parse v3.x.y` locally. If they differ, a moved tag was force-pushed and the runner used stale cache. Workflows now `git fetch --tags --force`, but verify.
3. **Container is running the new image (not a recycled old one).**
```bash
ssh root@staging.golfcards.club 'docker ps --format "{{.Names}} {{.Status}}"; docker inspect golfgame-app-1 --format "{{.Created}}"'
```
`Up X seconds/minutes` with a recent `.Created` time; not `Up 13 hours`. A compose restart picks up the *current* `:latest` image — confirm that image was built by this release, not a prior one.
4. **New code is actually in the container.** Introspect a signature/attribute that changed in this release:
```bash
ssh root@staging.golfcards.club 'docker exec golfgame-app-1 python -c "
import sys, inspect; sys.path.insert(0, \"/app/server\")
from services.game_logger import GameLogger
print(inspect.signature(GameLogger.log_game_start_async).parameters.keys())
"'
```
5. **Container env has every var the code reads.** Any config added to `server/config.py` this release needs TWO edits to flow through: `.env` on the host AND `- FOO=${FOO:-default}` in the compose yaml's `environment:` block. Setting only `.env` silently does nothing.
```bash
ssh root@staging.golfcards.club 'docker exec golfgame-app-1 printenv | grep -iE "YOUR_NEW_VAR"'
```
6. **DB schema + invariants hold.** Sample the tables this release touches and confirm the new columns/values look right:
```bash
ssh root@staging.golfcards.club 'docker exec golfgame-postgres-1 psql -U golf -d golf -c "SELECT status, COUNT(*) FROM games_v2 GROUP BY status;"'
```
7. **End-to-end smoke.** For a feature visible through the API, curl it and verify the response shape and content match expectations:
```bash
curl -s 'https://staging.golfcards.club/api/stats/leaderboard?metric=wins' | python3 -m json.tool
```
For features that only fire on specific game events (GAME_OVER, abandonment, etc.), run a soak game or manual repro and re-check the DB — don't assume "code is deployed" = "code has executed."
**If any step fails, stop and diagnose before running the next release.** Cascading hotfixes amplify the problem — each force-moved tag is another chance for the runner's cache to lie to you.
## Common Development Tasks
### Adjusting Animation Speed
Edit `timing-config.js` - all timings are centralized there.
### Adding New Animations
1. Add method to `CardAnimations` class in `card-animations.js`
2. Use anime.js, not CSS transitions
3. Track in `activeAnimations` Map for cancellation support
4. Add timing config to `timing-config.js` if needed
### Debugging Animations
```javascript
// Check what's animating
console.log(window.cardAnimations.activeAnimations);
// Force cleanup
window.cardAnimations.cancelAll();
// Check timing config
console.log(window.TIMING);
```
### Testing CPU Behavior
Adjust delays in `server/ai.py` `CPU_TIMING` dict.
### Running Tests
```bash
# All server tests
cd server && pytest -v
# AI simulation
python server/simulate.py 500
```
## Important Patterns
### No CSS Transitions on Cards
Cards animate via anime.js only. The following should NOT have `transition` (especially on `transform`):
- `.card`, `.card-inner`
- `.real-card`, `.swap-card`
- `.held-card-floating`
Card hover effects are handled by `CardAnimations.hoverIn()/hoverOut()` methods.
CSS may still use box-shadow transitions for hover glow effects.
### State Differ Logic (triggerAnimationsForStateChange)
The state differ in `app.js` detects what changed between game states:
**STEP 1: Draw Detection**
- Detects when `drawn_card` goes from null to something
- Triggers draw animation (from deck or discard)
- Sets `isDrawAnimating` flag
**STEP 2: Discard/Swap Detection**
- Detects when `discard_top` changes and it was another player's turn
- Triggers swap or discard animation
- **Important:** Skip STEP 2 if STEP 1 detected a draw from discard (the discard change was from REMOVING a card, not adding one)
### Animation Overlays
Complex animations create temporary overlay elements:
1. Create `.draw-anim-card` positioned over source
2. Hide original card (or set `opacity: 0` on discard pile during draw-from-discard)
3. Animate overlay
4. Remove overlay, reveal updated card, restore visibility
### Fire-and-Forget for Opponents
Opponent animations don't block - no callbacks needed:
```javascript
cardAnimations.animateOpponentFlip(cardElement, cardData);
```
### Common Animation Pitfalls
**Card position before append:** Always set `left`/`top` styles BEFORE appending overlay cards to body, otherwise they flash at (0,0).
**Deal animation source:** Use `getDeckRect()` for deal animations, not `getDealerRect()`. The dealer rect returns the whole player area, causing cards to animate at wrong size.
**Element rects during hidden:** `visibility: hidden` still allows `getBoundingClientRect()` to work. `display: none` does not.
## Dependencies
### Server
- FastAPI + uvicorn (web framework & ASGI server)
- websockets (WebSocket support)
- asyncpg (PostgreSQL async driver)
- redis (state caching, pub/sub)
- bcrypt (password hashing)
- resend (email service)
- python-dotenv (environment management)
- sentry-sdk (error tracking, optional)
### Client
- anime.js (animations)
- No other frameworks
### Infrastructure
- PostgreSQL (event sourcing, auth, stats, game logs)
- Redis (state caching, pub/sub)
## Game Rules Reference
- 6 cards per player in 2x3 grid
- Lower score wins
- Matching columns cancel out (0 points)
- Jokers are -2 points
- Kings are 0 points
- Game ends when a player flips all cards