cloudflare

Guides development across Cloudflare Workers, storage, AI, networking, and security products.

Updated Oct 15, 2019
One-click install
npx skills add https://github.com/kkkaoru/dotfiles --skill cloudflare-kkkaoru
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: cloudflare
Source: https://github.com/kkkaoru/dotfiles/tree/main/.agents/skills-stroage/cloudflare
Command: npx skills add https://github.com/kkkaoru/dotfiles --skill cloudflare-kkkaoru

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Cloudflare's platform spans dozens of products with evolving APIs, limits, and configuration options, making it hard to pick the right product and write correct code without outdated assumptions. This Skill routes any Cloudflare task to the right product reference and biases toward retrieving live docs over stale pre-trained knowledge. ## Core Features & Use Cases - Decision trees for product selection: Quickly map needs like "run code", "store data", "feature flags", or "AI inference" to the right Cloudflare product (Workers, D1, R2, KV, Vectorize, Flagship, and more). - Progressive per-product references: Over 50 reference directories (workers, durable-objects, d1, workers-ai, terraform, wrangler, etc.) with README, configuration, API, patterns, and gotchas files loaded only as needed. - Retrieval-first guidance: Explicit instructions to verify limits, pricing, and type signatures against live Cloudflare docs, wrangler config schemas, and workers-types packages. - Use Case: You need to build a RAG chatbot on Cloudflare. The decision tree routes you to Workers AI for embeddings, Vectorize for the vector database, and the agents-sdk reference for stateful agent logic, with quick-start code for each. ## Quick Start Ask the agent to help you build a Cloudflare Worker that stores data in D1 and caches responses in KV, and it will load the relevant product references to guide the implementation.

Frequently Asked Questions about cloudflare

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I choose between Cloudflare Workers, Pages, and Durable Objects?

Use Workers for serverless edge functions, Pages for full-stack web apps with Git deploys, and Durable Objects for stateful coordination or real-time workloads. The skill's decision trees route based on whether you need compute, hosting, or strongly-consistent state.

What Cloudflare storage should I use for my application?

Use KV for high-read key-value config and sessions, D1 for relational SQLite data, R2 for S3-compatible object storage, and Durable Object storage for strongly-consistent per-entity state. The skill provides a full storage decision tree including Queues, Vectorize, and Hyperdrive.

How do I build a RAG application on Cloudflare Workers AI?

Combine Workers AI embedding models like bge-base-en-v1.5 with Vectorize as the vector database, then run LLM inference via the env.AI binding. The workers-ai and vectorize references include model selection guidance and integration patterns.

Can I manage Cloudflare infrastructure with Terraform or Pulumi?

Yes, the skill includes terraform and pulumi references covering provider setup, authentication with API tokens, resource configuration, and v5 breaking-change migration. It recommends provider-first management without mixing wrangler.jsonc for the same resources.

Why does the skill prefer fetching Cloudflare docs over built-in knowledge?

Cloudflare limits, pricing, API signatures, and compatibility flags change frequently, so pre-trained knowledge may be outdated. The skill instructs retrieval from developers.cloudflare.com, wrangler config schemas, and workers-types packages before citing specifics.

When should I use Smart Placement for a Cloudflare Worker?

Enable Smart Placement when a Worker with a fetch handler makes multiple backend calls to geographically concentrated infrastructure. Avoid it for static content, pure edge logic, or Pages assets with run_worker_first enabled.