Cloudflare Worker Deployer

There is no Cloudflare Workers skill on ClawHub. This is the first one.

Cloudflare Worker Deployer gives your agent the full Workers lifecycle from a single skill: scaffold a new project with correct TypeScript config, configure bindings (KV, D1, R2, Durable Objects, queues, cron), set secrets without exposing them, deploy to staging before production, tail logs filtered by status or method, roll back to a specific previous deployment by ID, and manage custom domain routes. Eight wrangler.toml templates cover the most common project patterns out of the box.

The skill enforces safety rules by default: confirm before every production deploy, show what will be destroyed before any destructive operation, never hardcode secrets in generated config. It pairs directly with API Monetizer (add x402 payments to deployed Workers) and Greyline: Endpoint Monitor (monitor deployed Workers for health).

::What This Skill Does::

Cloudflare Worker Deployer gives your agent structured procedures to operate the complete Cloudflare Workers stack using the wrangler CLI. Your agent learns to scaffold new projects, configure resource bindings, manage secrets, deploy to environments, tail logs, roll back bad deploys, and inspect performance data — all without you navigating dashboards or memorizing wrangler command syntax.

The skill covers every major Cloudflare Workers primitive: KV namespaces, R2 object storage, D1 SQLite databases, Durable Objects, Queues, AI bindings, service bindings, cron triggers, and custom domain routing. It includes ready-to-use wrangler.toml templates for common project configurations and enforces a safety protocol around destructive operations: production deploys, rollbacks, and data deletions always require your explicit confirmation.

This is the only Cloudflare skill of its kind on ClawHub. There is no equivalent to the popular Vercel skill for Workers operators. This fills that gap.

::Use Cases::

Shipping a new API Worker: Scaffold the project, wire up D1 and KV, set secrets, deploy to staging for testing, then promote to production on the same domain route. No context-switching, no terminal spelunking.

Debugging a live Worker: Tail real-time logs filtered by error status, correlate exception messages with recent deploys, roll back to the last clean deployment ID if needed, and confirm the fix held.

Managing storage infrastructure: Create KV namespaces for session caching, R2 buckets for asset storage, or D1 databases for structured data. Run migrations against local dev before applying to production. Inspect and modify key-value pairs directly from chat.

Monitoring production performance: Query the Cloudflare Analytics GraphQL API for request volume, error rates, CPU time distribution (P50/P99), and duration metrics across any time window.

Setting up scheduled jobs: Configure cron triggers in wrangler.toml, export the scheduled handler, test locally with --test-scheduled, and deploy. The agent handles the syntax; you define the schedule.

::Who It's For::

Engineers deploying backend logic, APIs, or scheduled jobs on Cloudflare Workers who want to operate infrastructure from chat rather than switching between terminal and dashboard.

Founders and solo operators running Workers-based products who need to deploy, debug, and maintain infrastructure without deep wrangler expertise.

Teams using Workers as the backbone for SaaS products, edge APIs, or data pipelines who want a shared agent capable of executing standard deployment procedures.

Anyone already using the Vercel skill who also has Workers infrastructure. This skill complements it, covering the Cloudflare side of a hybrid edge stack.