cloudflare

An agent skill by cloudflare, from cloudflare/skills. Tags: architecture, cloud, cloudflare, product-strategy, security.

What it does

Discover and choose Cloudflare products for apps, APIs, AI agents, storage, networking, and security. Use for architecture and product selection, including when the user describes a need without naming a Cloudflare product; then find the relevant skill or documentation.

Install

With the skills CLI, which installs into Claude Code, Codex, Cursor and other agents:

npx skills add cloudflare/skills --skill cloudflare

Or copy the skill folder into Claude Code's skills directory by hand (~/.claude/skills for every project, or .claude/skills inside one):

git clone --depth 1 https://github.com/cloudflare/skills
cp -r skills/skills/cloudflare ~/.claude/skills/cloudflare

Safety box score

Grade C, moderate risk: 33 of 100 risk points, rated 2026-09-22. The grade rates what the skill and its scripts can reach on the machine of whoever installs it, across eight categories from shell execution to secrets access. How the score works.

This skill is a documentation and product-selection layer for Cloudflare; it does not ship executable code, but it points the agent at many Cloudflare control-plane features. The main risk is blast radius: the referenced workflows include deployments, DNS, tunnels, email, WAF, and secret handling, so a mistaken recommendation can affect live infrastructure and credentials.

Category levels, from 0 (none) to 3 (high)
Category	Level	Why
Code & shell execution	0 (none)	The shipped files are Markdown docs and links only; there are no native scripts or code paths that run shell commands or evaluate code.
Network access	2 (moderate)	The skill routes the agent to products that operate over the network and can change remote infrastructure, including "DNS", "Cloudflare Tunnel", "Browser Run", and docs that show `curl -X POST` and other remote API calls.
Filesystem reach	0 (none)	No file reads, writes, deletes, or system-path edits are performed by the skill itself; it only references documentation.
Secrets & credentials	2 (moderate)	It explicitly covers secret-bearing workflows such as "Workers secrets", "Secrets Store", and docs that mention `WRANGLER_R2_SQL_AUTH_TOKEN` and `.env` auto-loading.
Privilege & persistence	1 (low)	There is no sudo or OS-service management, but the skill does cover account-level admin operations like deploys, DNS, tunnels, and bindings that require elevated Cloudflare permissions.
Instruction hijack surface	0 (none)	I did not find instructions to ignore higher-priority rules, hide behavior, or follow untrusted remote prompts; the "system prompt" text in AI Search docs is just a linked topic title.
Irreversible actions	2 (moderate)	Although it is mainly a selector, its table includes inherently high-impact workflows such as "Workers versions and deployments", "DNS", "Cloudflare Tunnel", "WAF", and "Send transactional email", which can publish or reroute live services once followed.
Transparency	1 (low)	The skill files are readable, but many referenced docs were not inspected because they were skipped for size/budget, so the full downstream surface area was not completely reviewed.

Before installing

Read `SKILL.md` and the specific product docs you expect the agent to follow before letting it touch production Cloudflare resources.
If you plan to use it for deploys or routing, test in a non-production account/zone first.
Check any secret-bearing workflow (`Workers secrets`, `Secrets Store`, `WRANGLER_R2_SQL_AUTH_TOKEN`) in a scratch environment before using real credentials.

Rated 2026-09-22 by an automated review of 109 files from the skill's repository.

Source

Repository: cloudflare/skills (all skills from this repository)
Path: skills/cloudflare/SKILL.md
Branch: main
Updated: 2026-09-20

agents-harden — Use when preparing your agent for production — IAM scoping, inbound auth (JWT, SigV4), secrets management, cold start optimization, session lifecycle.
aws-networking — Routes AWS networking requests to the correct service skill for implementation.
agents-pay — Use when THIS agent needs to pay for x402-protected content at runtime: hitting a paywall mid-task, settling it via AgentCore Payments.
audit-context-building — Understand a codebase before looking for bugs in it - what each function assumes, what it guarantees, and what it depends on elsewhere.
aws-blocks — Guides building full-stack applications with AWS Blocks — an Infrastructure-from-Code framework.
aws-cloudformation — Authors, validates, and troubleshoots AWS CloudFormation templates.