Cloudflare Workers Deploy

## Who Is This For

Developers deploying to Cloudflare Workers and Pages. This plugin covers the full lifecycle — from designing an edge-optimised architecture to deploying with wrangler, tracking deployments, and following production best practices.

## 3-Step Quickstart

1. **Design your edge deployment** — ask: `"Use edge-computing-patterns to suggest the best Cloudflare Workers architecture for a globally distributed API"`. The skill covers Worker routing, KV storage, and D1 database patterns.
2. **Deploy with wrangler** — ask: `"Use wrangler to deploy my Worker in ./src/worker.ts to production"`. The wrangler and dicklesworthstone-wrangler skills handle the Wrangler CLI, KV namespaces, R2 buckets, and secrets.
3. **Track and manage** — ask: `"Use deployment-management to list all active Workers deployments and check their health status"`. For a complete deployment strategy, ask: `"Use the Deployment Engineer to create a deployment plan for rolling out v2 with zero downtime"`.

## Step → Skill → Input → Output

| Step | Skill | What You Provide | What You Get |
|------|-------|------------------|--------------|
| Architecture design | edge-computing-patterns | App requirements | Workers architecture + routing recommendations |
| Deploy | wrangler / dicklesworthstone-wrangler | Worker source path + environment | Deploy URL + wrangler output |
| Production deploy flow | deploying-to-production | Repo URL + target (Vercel/CF) | GitHub repo + live deployment |
| Track deployments | deployment-management | Infrastructure name / range | Deployment status + health report |
| Deployment strategy | Deployment Engineer | App type + rollout requirements | Step-by-step deployment plan |

## Common Failures & Fixes

- **wrangler fails with "not authenticated"** → Run `wrangler login` in your terminal first, or set `CLOUDFLARE_API_TOKEN` as an env var.
- **edge-computing-patterns gives a generic answer** → Specify your use case: `"globally distributed read-heavy API with 50ms SLA"` gives a much more targeted pattern.
- **deployment-management shows no deployments** → Check that the infrastructure range or project name matches exactly what's in your Cloudflare dashboard.

## Success Criteria

You know it worked when:
- edge-computing-patterns returns a concrete architecture diagram or routing config
- wrangler or dicklesworthstone-wrangler outputs a `workers.dev` or custom domain URL
- deployment-management confirms all Worker instances are healthy