@arthai/agents 1.0.0

package/dist/plugins/prime/skills/setup/SKILL.md

@@ -0,0 +1,691 @@
+---
+name: setup
+description: "Bootstrap any project from empty repo to deployed app. Modular — run full or target a module. Usage: /setup <module>"
+user-invocable: true
+arguments: "<module>"
+---
+
+# /setup — Project Bootstrap
+
+Universal project setup — from fresh repo to deployed app. Modular design:
+run the full sequence or target a single module. Adapts to any stack by
+discovering the project from CLAUDE.md, package files, and docker-compose.
+
+Every module is idempotent: re-run safely to fix or verify.
+
+## Argument Parsing
+
+| Input | Action |
+|-------|--------|
+| `/setup` | Full interactive setup — detect what's needed, skip what's done |
+| `/setup status` | Show readiness table across all modules |
+| `/setup prerequisites` | System requirements + account creation |
+| `/setup repo` | Git + project structure |
+| `/setup deps` | Dependencies (venv, npm, cargo, etc.) |
+| `/setup infra` | Docker services (DB, cache, queue) |
+| `/setup database` | Migrations + seed data |
+| `/setup credentials` | API keys + env files |
+| `/setup profiles` | Bot/service identity + social presence |
+| `/setup verify-local` | Start app + smoke tests |
+| `/setup claude-agents` | Toolkit + CLAUDE.md |
+| `/setup arth` | Arth dashboard registration |
+| `/setup ci` | GitHub Actions workflow |
+| `/setup discord` | Discord server + bot + webhooks + ChatOps |
+| `/setup secrets-sync` | Push secrets to GitHub + cloud provider |
+| `/setup deploy` | Cloud deployment (Railway, Vercel, AWS, etc.) |
+| `/setup verify-prod` | Production smoke tests |
+| `/setup monitoring` | Sentry / error tracking |
+
+## Project Context Discovery
+
+Before running any module, detect the project's stack:
+
+1. **Read CLAUDE.md** for tech stack, services, ports, test commands
+2. **Scan project files** if CLAUDE.md is incomplete:
+
+| File | Detects |
+|------|---------|
+| `package.json` | Node/frontend framework (Next.js, Vite, React) |
+| `requirements.txt` / `pyproject.toml` | Python framework (FastAPI, Django, Flask) |
+| `go.mod` | Go backend |
+| `Cargo.toml` | Rust backend |
+| `docker-compose.yml` | Infrastructure services (DB, Redis, etc.) |
+| `.env.example` / `.env.local` / `.env` | Required credentials |
+| `alembic/` / `prisma/` / `migrations/` | Migration tool |
+| `Dockerfile` / `Procfile` / `railway.json` | Deploy target |
+| `backend/` + `frontend/` | Full-stack layout |
+| `.claude/plans/` | Existing project plans with architecture |
+
+3. **Determine env file convention**: Check config/settings for `env_file` path.
+   Common patterns: `.env`, `.env.local`, `backend/.env`, `.env.development`
+
+## Module Overview
+
+```
+ 0. Prerequisites ─── accounts + software
+ 1. Repo Structure ── git init + .gitignore + scaffolding
+ 2. Dependencies ──── language-specific package install
+ 3. Infrastructure ── Docker (DB + cache + queue)
+ 4. Database ──────── migrations + seed data
+ 5. Credentials ───── API keys + env files
+ 6. Profiles ──────── Bot/service identity (if applicable)
+ 7. Local Verify ──── start app + smoke tests
+ 8. Claude Agents ─── toolkit + CLAUDE.md
+ 9. Arth ──────────── dashboard registration
+10. GitHub CI ─────── lint + test on every PR
+11. Discord ───────── notifications + ChatOps
+12. Secrets Sync ──── push to GitHub + cloud
+13. Deploy ────────── cloud deployment
+14. Prod Verify ───── health + smoke tests
+15. Monitoring ────── error tracking (optional)
+```
+
+For `/setup` (full), walk through each module in order, skipping any already complete.
+For `/setup <module>`, jump directly to that module.
+
+---
+
+## Module 0: Prerequisites
+
+Detect what's needed from the project stack, then check + guide.
+
+### Software Detection
+
+| If project has... | Require |
+|-------------------|---------|
+| `requirements.txt` | Python 3.11+ (`python3.11 --version`) |
+| `package.json` | Node 20+ (`node --version`) |
+| `go.mod` | Go 1.21+ (`go version`) |
+| `Cargo.toml` | Rust (`rustc --version`) |
+| `docker-compose.yml` | Docker (`docker --version`) |
+| Any git repo | Git (`git --version`) |
+| GitHub remote | GitHub CLI (`gh --version`) |
+| Railway deploy target | Railway CLI (`railway --version`) |
+
+### Account Detection
+
+Scan `.env.example` or config/settings for required services:
+
+| Env var pattern | Account needed | URL |
+|----------------|---------------|-----|
+| `ANTHROPIC_API_KEY` | Anthropic | console.anthropic.com |
+| `OPENAI_API_KEY` | OpenAI | platform.openai.com |
+| `X_BEARER_TOKEN` / `TWITTER_*` | X/Twitter Developer | developer.x.com |
+| `TELEGRAM_BOT_TOKEN` | Telegram (@BotFather) | telegram.org |
+| `STYTCH_*` | Stytch | stytch.com |
+| `LIVEKIT_*` | LiveKit | cloud.livekit.io |
+| `DEEPGRAM_API_KEY` | Deepgram | console.deepgram.com |
+| `STRIPE_*` | Stripe | dashboard.stripe.com |
+| `SENTRY_DSN` | Sentry | sentry.io |
+| `DISCORD_*` | Discord | discord.com |
+| `RAILWAY_TOKEN` | Railway | railway.app |
+
+Show status table of what's installed/missing.
+
+---
+
+## Module 1: Repo Structure
+
+**Check:** `.git/` exists, `.gitignore` covers secrets and build artifacts.
+
+**Actions:**
+```bash
+git init                          # if no .git/
+```
+
+**Standard .gitignore entries** (append if missing):
+```
+.env .env.local .env.*.local
+__pycache__/ *.pyc .pytest_cache/
+node_modules/ .next/ dist/ build/
+.venv/ venv/
+*.db *.sqlite3
+.DS_Store
+```
+
+**Scaffolding** — detect from CLAUDE.md or project plan, or ask user:
+
+| Stack | Structure |
+|-------|-----------|
+| FastAPI | `backend/app/{core,models,services,workers,api}/`, `backend/tests/`, `backend/alembic/` |
+| Django | `backend/<app>/{models,views,serializers,tests}/` |
+| Next.js | `frontend/src/{app,components,lib}/` |
+| Full-stack | both + `docker-compose.yml` |
+| Go | `cmd/`, `internal/`, `pkg/` |
+| Rust | `src/`, `tests/` |
+
+---
+
+## Module 2: Dependencies
+
+Detect stack and install:
+
+### Python
+```bash
+cd backend && python3.11 -m venv .venv && source .venv/bin/activate
+pip install -r requirements.txt
+pip install greenlet 2>/dev/null   # often missing, needed by SQLAlchemy async
+```
+
+### Node
+```bash
+cd frontend && npm install
+```
+
+### Go
+```bash
+go mod download
+```
+
+### Rust
+```bash
+cargo build
+```
+
+### ML Models
+
+If `requirements.txt` contains ML packages, pre-download models:
+
+| Package | Model to pre-download |
+|---------|----------------------|
+| `sentence-transformers` | `python3 -c "from sentence_transformers import SentenceTransformer; SentenceTransformer('<model-name>')"` |
+| `fasttext-langdetect` | `python3 -c "from ftlangdetect import detect; detect('hello')"` |
+| `transformers` | Check code for `from_pretrained()` calls |
+
+### Validate
+
+```bash
+# Python
+source backend/.venv/bin/activate && python3 -c "import app; print('OK')"
+# Node
+cd frontend && npm run build
+```
+
+---
+
+## Module 3: Infrastructure (Docker)
+
+**Check:** `docker-compose.yml` exists, services are running.
+
+Detect services from docker-compose.yml and verify each:
+
+| Service | Health Check |
+|---------|-------------|
+| PostgreSQL | `docker compose exec <svc> pg_isready` |
+| Redis | `docker compose exec <svc> redis-cli ping` |
+| MySQL | `docker compose exec <svc> mysqladmin ping` |
+| MongoDB | `docker compose exec <svc> mongosh --eval "db.adminCommand('ping')"` |
+
+**Start:**
+```bash
+docker compose up -d && docker compose ps
+```
+
+---
+
+## Module 4: Database Setup
+
+Detect migration tool and run:
+
+| Tool | Detect | Run |
+|------|--------|-----|
+| Alembic | `alembic/` dir | `alembic upgrade head` |
+| Prisma | `prisma/` dir | `npx prisma migrate deploy` |
+| Django | `manage.py` | `python manage.py migrate` |
+| Knex | `knexfile.*` | `npx knex migrate:latest` |
+| Drizzle | `drizzle/` dir | `npx drizzle-kit push` |
+
+**Seed data:** Look for seed scripts:
+- `scripts/seed*.py` → `python -m scripts.<name>`
+- `prisma/seed.ts` → `npx prisma db seed`
+- `manage.py` → `python manage.py loaddata`
+- `seeds/` directory
+
+**Verify:** Query row counts on key tables.
+
+---
+
+## Module 5: Credentials
+
+### Workflow
+
+1. Detect env file convention from project config/settings
+2. Create from template: `cp -n .env.example <env-file>`
+3. Auto-generate what we can (secret keys, webhook secrets):
+   ```bash
+   python3 -c "import secrets; print(secrets.token_urlsafe(32))"
+   ```
+4. Check for placeholders: grep for `=...`, `=sk-ant-...`, `=AAAA...`, `=change-me`
+5. Show status table (per credential)
+6. Guide user through each missing credential
+7. Validate each with API call where possible
+8. Final check: import settings/config to verify all loads
+
+**IMPORTANT:** Never store credentials yourself. Ask the user to paste values. Never echo credentials back.
+
+### Universal Credential Guides
+
+Pick the guides relevant to the project's `.env.example`:
+
+#### Anthropic API Key
+- **Get:** console.anthropic.com → API Keys → Create Key (starts with `sk-ant-`)
+- **Cost:** Pay-per-use (~$30/mo at 10K requests)
+- **Validate:**
+```bash
+curl -s -H "x-api-key: $ANTHROPIC_API_KEY" -H "anthropic-version: 2023-06-01" \
+  -H "content-type: application/json" \
+  https://api.anthropic.com/v1/messages \
+  -d '{"model":"claude-sonnet-4-5","max_tokens":1,"messages":[{"role":"user","content":"hi"}]}' \
+  | python3 -c "import sys,json; r=json.load(sys.stdin); print('OK' if 'content' in r else f'FAIL: {r}')"
+```
+
+#### OpenAI API Key
+- **Get:** platform.openai.com → API Keys → Create Key (starts with `sk-`)
+- **Cost:** Pay-per-use
+- **Var:** `OPENAI_API_KEY`
+- **Validate:**
+```bash
+curl -s -H "Authorization: Bearer $OPENAI_API_KEY" \
+  https://api.openai.com/v1/models \
+  | python3 -c "import sys,json; r=json.load(sys.stdin); print('OK' if 'data' in r else f'FAIL: {r}')"
+```
+
+#### Twitter/X API (6 Keys)
+
+**Step-by-step:**
+1. Go to developer.x.com → sign in → accept developer agreement
+2. Create Project + App inside it
+3. App → Settings → User authentication settings → Set **Read and Write** permissions
+4. App → Keys and Tokens:
+   - Consumer Keys → Regenerate → `X_API_KEY`, `X_API_SECRET`
+   - Bearer Token → Regenerate → `X_BEARER_TOKEN`
+   - Access Token → Generate (AFTER setting permissions) → `X_ACCESS_TOKEN`, `X_ACCESS_SECRET`
+5. Get user ID:
+```bash
+curl -s -H "Authorization: Bearer $X_BEARER_TOKEN" \
+  "https://api.x.com/2/users/by/username/$X_BOT_USERNAME" \
+  | python3 -c "import sys,json; print(json.load(sys.stdin)['data']['id'])"
+```
+6. Purchase API credits at developer.x.com → Credits ($5-$500, pay-per-use, never expire)
+
+**Validate:**
+```bash
+curl -s -H "Authorization: Bearer $X_BEARER_TOKEN" "https://api.x.com/2/users/me" \
+  | python3 -c "import sys,json; r=json.load(sys.stdin); print(f'OK: @{r[\"data\"][\"username\"]}' if 'data' in r else f'FAIL: {r}')"
+```
+
+#### Telegram Bot Token (Free)
+- **Get:** Telegram → @BotFather → `/newbot` → choose name + username (must end in `bot`)
+- **Validate:**
+```bash
+curl -s "https://api.telegram.org/bot${TELEGRAM_BOT_TOKEN}/getMe" \
+  | python3 -c "import sys,json; r=json.load(sys.stdin); print(f'OK: @{r[\"result\"][\"username\"]}' if r.get('ok') else f'FAIL: {r}')"
+```
+
+#### Stytch Auth
+- **Get:** stytch.com → Project → API Keys
+- **Vars:** `STYTCH_PROJECT_ID`, `STYTCH_SECRET`, `NEXT_PUBLIC_STYTCH_PUBLIC_TOKEN`
+- **Validate:** Import Stytch client and call a test endpoint
+
+#### LiveKit (Real-time Voice/Video)
+- **Get:** cloud.livekit.io → Project → Keys
+- **Vars:** `LIVEKIT_URL`, `LIVEKIT_API_KEY`, `LIVEKIT_API_SECRET`
+- **Validate:**
+```bash
+curl -s "$LIVEKIT_URL" && echo "OK"
+```
+
+#### Deepgram (Speech-to-Text)
+- **Get:** console.deepgram.com → API Keys
+- **Var:** `DEEPGRAM_API_KEY`
+- **Validate:**
+```bash
+curl -s -H "Authorization: Token $DEEPGRAM_API_KEY" \
+  https://api.deepgram.com/v1/projects \
+  | python3 -c "import sys,json; r=json.load(sys.stdin); print('OK' if 'projects' in r else f'FAIL: {r}')"
+```
+
+#### Stripe
+- **Get:** dashboard.stripe.com → Developers → API Keys
+- **Vars:** `STRIPE_SECRET_KEY`, `STRIPE_PUBLISHABLE_KEY`, `STRIPE_WEBHOOK_SECRET`
+- **Validate:**
+```bash
+curl -s -u "$STRIPE_SECRET_KEY:" https://api.stripe.com/v1/balance \
+  | python3 -c "import sys,json; r=json.load(sys.stdin); print('OK' if 'available' in r else f'FAIL: {r}')"
+```
+
+#### Sentry (Optional)
+- **Get:** sentry.io → Create project → copy DSN
+- **Var:** `SENTRY_DSN=https://...@sentry.io/...`
+- **Cost:** Free tier (5K errors/mo)
+
+### Final Config Validation
+
+Attempt to import the project's settings/config module to verify all env vars load:
+
+```bash
+# Python (Pydantic Settings)
+python3 -c "from app.core.config import settings; print('Config OK')"
+
+# Node (dotenv)
+node -e "require('dotenv').config(); console.log('Config OK')"
+```
+
+---
+
+## Module 6: Bot/Service Profiles (If Applicable)
+
+Only run if the project includes bot/service integrations. Detect from env vars.
+
+### Twitter/X Profile
+- Display name, bio (160 chars), avatar, header image
+- Username change (note: rate limited, 24-48h cooldown on 429)
+- Follow relevant accounts for the project's domain
+
+### Telegram Bot Profile (via @BotFather)
+- `/setdescription` — bot description
+- `/setabouttext` — about text
+- `/setcommands` — command menu
+- `/setuserpic` — profile photo
+
+### Discord Bot Profile
+- Display name, avatar via Developer Portal
+
+---
+
+## Module 7: Local Startup & Verification
+
+Detect start commands from CLAUDE.md or project files:
+
+| Stack | Start Command |
+|-------|--------------|
+| FastAPI | `uvicorn app.main:app --reload --port 8000` |
+| Django | `python manage.py runserver` |
+| Next.js | `npm run dev` |
+| Vite | `npm run dev` |
+| Go | `go run cmd/server/main.go` |
+
+**Health checks:** Hit health endpoints from CLAUDE.md or common patterns (`/health`, `/api/health`, `/healthz`).
+
+**Tests:**
+
+| Stack | Test Command |
+|-------|-------------|
+| Python | `pytest --cov -v` |
+| Node | `npm test` |
+| Go | `go test ./...` |
+| Rust | `cargo test` |
+
+---
+
+## Module 8: Claude Agents + CLAUDE.md
+
+**Check:** `.claude/` directory exists, `CLAUDE.md` has no `<!-- TODO -->` placeholders.
+
+**First-time install:**
+```bash
+~/.claude-agents/install.sh --init .
+```
+
+**Existing project refresh:**
+```bash
+~/.claude-agents/install.sh .
+```
+
+**Populate CLAUDE.md:** Run `/scan` to auto-detect stack, ports, test commands, services.
+
+**Verify:** Grep CLAUDE.md for remaining `<!-- TODO -->` — fill any critical sections manually.
+
+---
+
+## Module 9: Arth Registration
+
+**Check:** `.arth.json` exists in project root.
+
+**Register:** Run `/arth add`.
+
+**Verify:** Visit http://localhost:3333, confirm project appears.
+
+---
+
+## Module 10: GitHub CI Workflows
+
+**Check:** `.github/workflows/ci.yml` exists.
+
+Generate CI workflow based on detected stack. The template should include:
+
+1. **Services** — mirror docker-compose.yml services as GitHub Actions service containers
+2. **Install** — language-specific dependency install
+3. **Lint** — project's lint command
+4. **Type check** — if applicable (mypy, tsc)
+5. **Test** — project's test command with coverage
+6. **Env vars** — fake values for all external API keys (tests should mock these)
+
+Detect test env vars from `.env.example` and generate fake values.
+
+---
+
+## Module 11: Discord (CI/Deploy Tracking + ChatOps)
+
+### 11a. Create Discord Server + Channels
+
+Recommended channels:
+- `#deployments` — automated CI/CD + deploy notifications (webhook)
+- `#claude-ops` — ChatOps commands for remote Claude Code control
+- `#general` — manual posts, PR summaries, discussions
+- `#alerts` — production errors, health check failures
+
+### 11b. Create Discord Bot
+
+1. discord.com/developers/applications → New Application
+2. Bot tab → Add Bot → copy token
+3. Enable Message Content Intent
+4. OAuth2 → URL Generator → scope: `bot` → permissions: Send Messages, Read History, Embed Links
+5. Invite to server
+
+### 11c. GitHub Webhook → #deployments
+
+1. Discord: right-click `#deployments` → Integrations → Webhooks → New Webhook → copy URL
+2. GitHub: repo → Settings → Webhooks → Add webhook
+   - URL: `<discord-webhook-url>/github`
+   - Events: Pull requests, Push, Workflow runs, Deployments, Issues
+
+### 11d. ChatOps via discord-mcp
+
+Configure `~/discord-mcp/.env`:
+```env
+DISCORD_TOKEN=<bot token>
+DISCORD_AUTHORIZED_USER_IDS=<your Discord user ID>
+DISCORD_OPS_CHANNEL=claude-ops
+CLAUDE_WORKING_DIR=<project-root>
+CLAUDE_PATH=claude
+```
+
+Start: `cd ~/discord-mcp && node build/index.js`
+
+Key commands: `fix ci`, `ci status`, `run tests`, `git status`, `pr status`, `check prod health`
+
+---
+
+## Module 12: Secrets Sync
+
+After all credentials are configured locally, sync to GitHub Secrets and cloud provider.
+
+### Detect What to Sync
+
+Read env file and identify real values (skip placeholders like `=...`, `=AAAA...`, `=change-me`).
+
+**Rules:**
+- Only sync secrets with real values — skip placeholders
+- Never sync local-only values (`DATABASE_URL`, `REDIS_URL`) — cloud providers auto-set these
+- Always confirm with user before pushing to remote platforms
+- Show confirmation table after syncing
+
+### GitHub Secrets
+```bash
+for var in <SECRET_LIST>; do
+  value=$(grep "^${var}=" <env-file> | cut -d= -f2-)
+  if [ -n "$value" ] && [ "$value" != "..." ]; then
+    echo "$value" | gh secret set "$var" --repo OWNER/REPO
+  fi
+done
+gh secret list --repo OWNER/REPO
+```
+
+### Cloud Provider
+```bash
+# Railway
+railway variables set VAR_NAME="$VALUE"
+
+# Vercel
+vercel env add VAR_NAME production < <(echo "$VALUE")
+
+# AWS
+aws ssm put-parameter --name "/app/VAR_NAME" --value "$VALUE" --type SecureString
+```
+
+---
+
+## Module 13: Cloud Deployment
+
+Detect deploy target from project files:
+
+| Signal | Platform | Setup |
+|--------|----------|-------|
+| `railway.json` / `Procfile` | Railway | `railway login && railway init && railway up` |
+| `vercel.json` / `next.config.*` | Vercel | `vercel && vercel --prod` |
+| `Dockerfile` + no platform | Railway or Fly.io | Ask user preference |
+| `serverless.yml` | AWS Serverless | `serverless deploy` |
+| `app.yaml` | Google Cloud | `gcloud app deploy` |
+
+### Railway (Common)
+1. `railway login && railway init`
+2. Add plugins: `railway add --plugin postgresql && railway add --plugin redis`
+3. Set env vars (Module 12)
+4. Deploy: `railway up`
+5. Get URL: `railway domain`
+
+### Post-Deploy
+- Run migrations on production
+- Seed data if first deploy
+- Register webhooks (Telegram, Stripe, etc.) with production URL
+
+---
+
+## Module 14: Production Verification
+
+After deploying:
+
+1. **Health check:** `curl -s $PROD_URL/health`
+2. **End-to-end tests:** Test core user flows against production
+3. **Check logs:** `railway logs` / cloud provider log viewer
+4. **Verify webhooks:** Send test events, confirm they arrive
+
+---
+
+## Module 15: Monitoring (Optional)
+
+### Sentry
+- sentry.io → Create project for your language → copy DSN
+- Set `SENTRY_DSN` in cloud provider env vars
+- Free tier: 5K errors/mo
+
+---
+
+## Status Check (`/setup status`)
+
+Show a readiness table:
+
+```
+| #  | Module            | Status | Notes                    |
+|----|-------------------|--------|--------------------------|
+| 0  | Prerequisites     | ✅/❌  | software + accounts      |
+| 1  | Repo              | ✅/❌  | git initialized          |
+| 2  | Dependencies      | ✅/❌  | packages installed       |
+| 3  | Infrastructure    | ✅/❌  | Docker services running  |
+| 4  | Database          | ✅/❌  | migrations + seed data   |
+| 5  | Credentials       | ✅/⚠️  | X of Y configured        |
+| 6  | Profiles          | ✅/❌  | bot profiles set up      |
+| 7  | Local Verify      | ✅/❌  | app starts, tests pass   |
+| 8  | Claude Agents     | ✅/❌  | toolkit installed         |
+| 9  | Arth              | ✅/❌  | registered               |
+| 10 | GitHub CI         | ✅/❌  | ci.yml exists            |
+| 11 | Discord           | ✅/❌  | server + bot + webhooks  |
+| 12 | Secrets Sync      | ✅/❌  | GitHub + cloud synced    |
+| 13 | Deploy            | ✅/❌  | deployed + healthy       |
+| 14 | Prod Verify       | ✅/❌  | smoke tests pass         |
+| 15 | Monitoring        | ⏭️     | optional                 |
+```
+
+### How to Check Each
+
+```bash
+# 0. Prerequisites
+python3.11 --version 2>/dev/null; node --version 2>/dev/null; docker --version 2>/dev/null; gh --version 2>/dev/null
+
+# 1. Repo
+test -d .git && echo "OK"
+
+# 2. Dependencies
+test -d backend/.venv 2>/dev/null || test -d node_modules 2>/dev/null || test -d .venv 2>/dev/null
+
+# 3. Infrastructure
+docker compose ps --format json 2>/dev/null | python3 -c "import sys,json; [print(json.loads(l)['Service'], json.loads(l).get('State','?')) for l in sys.stdin]"
+
+# 4. Database — attempt to query a table
+# (project-specific — detect from migrations)
+
+# 5. Credentials
+grep -cE '=\.\.\.|=change-me|=AAAA\.\.\.|=sk-ant-\.\.\.' <env-file> 2>/dev/null
+
+# 6. Profiles — manual check
+
+# 7. Local Verify
+curl -s http://localhost:<port>/health 2>/dev/null
+
+# 8. Claude Agents
+ls -la .claude/agents/ 2>/dev/null | wc -l
+
+# 9. Arth
+test -f .arth.json
+
+# 10. GitHub CI
+test -f .github/workflows/ci.yml
+
+# 11. Discord
+test -d ~/discord-mcp && test -f ~/discord-mcp/.env
+
+# 12. Secrets Sync
+gh secret list --repo OWNER/REPO 2>/dev/null | wc -l
+
+# 13. Deploy
+railway status 2>/dev/null || vercel ls 2>/dev/null
+
+# 14. Prod Verify
+curl -s $PROD_URL/health 2>/dev/null
+
+# 15. Monitoring
+grep -q 'SENTRY_DSN=https' <env-file>
+```
+
+---
+
+## Common Issues
+
+| Symptom | Cause | Fix |
+|---------|-------|-----|
+| `ValidationError` on startup | Missing env vars | `/setup credentials` |
+| `ConnectionRefusedError` :5432 | Postgres not running | `docker compose up -d` |
+| `ConnectionRefusedError` :6379 | Redis not running | `docker compose up -d` |
+| `ModuleNotFoundError: greenlet` | Missing SQLAlchemy async dep | `pip install greenlet` |
+| ML model download hangs | Large model on slow network | Retry, model cached after first download |
+| `alembic upgrade` fails | Tables already exist | `alembic stamp head` then re-run |
+| X API 429 errors | Rate limited | Wait, check daily limits |
+| X username change 429 | Rate limited on settings endpoint | Wait 24-48h, do NOT retry |
+| Telegram webhook fails | Wrong URL or token | Re-register webhook with production URL |
+| `No module named 'app'` | Wrong working directory | Run from correct directory |
+| CI fails service health | Container not ready | Add health check options to workflow |
+| Deploy fails | Missing Procfile or env vars | Check cloud provider logs |
+| pgvector missing | Wrong Postgres image | Use `ankane/pgvector:latest` |
+| `npm ERR! ERESOLVE` | Dependency conflict | `npm install --legacy-peer-deps` |
+| `install.sh` permission denied | Not executable | `chmod +x ~/.claude-agents/install.sh` |