Codex skills for Cloudflare
Cloudflare skills cover Workers, Pages, D1, R2, KV, Queues, Durable Objects and the CDN and WAF in front of them. Codex pins compatibility dates, generates Env types with wrangler types, applies D1 migrations locally before remotely, acks queue messages one at a time, and waits for your approval before production deploys or destructive Durable Object migrations.
Start with these 3
- Cloudflare Workers Edge code that respects the isolate model Codex sets up Workers with a pinned compatibility date, bindings repeated per environment, and Env types generated by wrangler types instead of hand-written. Code avoids per-request globals and module-scope I/O, pushes post-response work into waitUntil, tests run in the Workers Vitest pool, and production deploys go through versioned gradual rollouts with a known rollback.
- Cloudflare Pages Headers and redirects that actually apply Codex deploys Pages sites from the correct build output, lints _headers and _redirects with a bundled script before each deploy, and limits Pages Functions to the routes that need them. It knows which rules skip Function responses, keeps preview URLs out of search indexes, and says when a new project belongs on Workers with static assets instead.
- Cloudflare D1 SQLite at the edge, migrated and indexed properly Codex manages D1 schemas through numbered migrations applied local, then staging, then production, writes every query with bound parameters, and uses batch() where other databases would use a transaction. It reads EXPLAIN QUERY PLAN and rows_read to cut cost, records a Time Travel bookmark before risky changes, and says when KV or Durable Objects fit better.
7 more Cloudflare skills
-
Cloudflare R2
Zero-egress storage with uploads locked down -
Cloudflare Durable Objects
One object per room, state that survives eviction -
Cloudflare KV and Queues
Stop treating KV like a database -
Edge Functions
Run at the edge only what belongs there -
CDN Caching
High hit ratios without leaking anyone's page -
Hono Edge API
One typed Hono app for Workers, Bun, Deno and Node -
WAF Rules
Count first, block later, except narrowly
General skills with Cloudflare guidance
Not built around Cloudflare, but they include steps or tools for it. Not part of the install-all command.
-
SvelteKit
Server loads, real forms, the right adapter -
Astro Sites
Static by default, islands only where they earn it -
Cron Scheduling
Scheduled jobs that run once, on time, visibly -
Serverless Postgres
Pooled at runtime, direct for migrations, branched per PR -
Connection Pooling
Fit every pool inside max_connections -
DNS Configuration
DNS changes checked at the source, not the cache -
Netlify Deploys
Redirects that fire and previews that stay isolated -
SST v3 Apps
TypeScript infrastructure with links, not hardcoded ARNs -
Image Optimization
Smaller images, same pixels, faster hero
Every Cloudflare skill in one command
Installs these 10 into ~/.agents/skills/. Add --project after -- to install into the current repository instead.
The installer is a 104-line POSIX shell script. It fetches the skills you name from codexskills.dev and unpacks each into its own folder in your skills directory. No sudo, no telemetry, and every skill page shows the full SKILL.md first. Read install.sh
curl -fsSL codexskills.dev/install.sh | sh -s -- cloudflare-workers cloudflare-pages cloudflare-d1 cloudflare-r2 cloudflare-durable-objects cloudflare-kv-queues edge-functions cdn-caching hono-edge-api waf-rules Cloudflare FAQ
Is there a Codex skill for Cloudflare Workers and Wrangler?
Cloudflare Workers configures wrangler.jsonc with a pinned compatibility_date and nodejs_compat, repeats bindings for every environment, and regenerates Env types with npx wrangler types after each binding change. Tests run inside the Workers runtime through @cloudflare/vitest-pool-workers, and production deploys use wrangler versions upload and gradual rollouts after you confirm. Hono Edge API covers routing and validation on top of a Worker.
Why are my D1 migrations not applied in production?
Cloudflare D1 names the usual cause: running wrangler d1 migrations apply without --remote, which updates only the local database. The skill applies numbered migrations locally, then to staging, then to production, uses batch() as the transaction primitive, and records a Time Travel bookmark before risky changes. It also reads EXPLAIN QUERY PLAN and rows_read to cut query cost.
Should I use Workers KV, D1 or Durable Objects?
Each skill says when another one fits better. Cloudflare KV and Queues treats KV as an eventually consistent, read-heavy cache where writes take up to about 60 seconds to propagate and each key takes one write per second. Cloudflare D1 is for relational data queried with SQL, and Cloudflare Durable Objects is for strongly consistent per-entity state, rate limiters and WebSocket rooms on the Hibernation API.
How do I install all 10 Cloudflare skills at once?
Pass every name to the installer in one command: curl -fsSL https://codexskills.dev/install.sh | sh -s -- cloudflare-workers cloudflare-pages cloudflare-d1 cloudflare-r2 cloudflare-durable-objects cloudflare-kv-queues edge-functions cdn-caching hono-edge-api waf-rules. Add --project right after -- to put them in the current repository's .agents/skills/ so your team gets them too.
Which Cloudflare skill should I try first?
Cloudflare Workers is the closest fit for most Cloudflare work. Mention it with $cloudflare-workers in a prompt to load it on purpose; after that, Codex picks it up on its own whenever a task matches.