cloudflare-access-mcp

cloudflare-access-mcp is a skill for Claude Code from Goodsmileduck/claude-registry. It costs 164 tokens per session (2,279 once invoked), scanned A, original, MIT.

Guidance for protecting a remote MCP server with Cloudflare OAuth or SSO, which lets approved users sign in before using it. It covers a hosted portal, a self-managed Cloudflare Access setup, and Terraform-based configuration.

In plain words
What is it for?
Use it to place OAuth in front of existing MCP URLs, configure a Cloudflare Access application, or create the setup as Terraform infrastructure.
Why use it?
It helps choose an authentication setup and avoid configuration mistakes when exposing an MCP server to users or other systems.

Skill for Claude Code

Written for Claude Code: shipped in a Claude Code plugin. Also seen: mentions Claude Code.

Part of the cloud-platform-skills plugin — 7 skills shipped together

Good fit Use it to place OAuth in front of existing MCP URLs, configure a Cloudflare Access application, or create the setup as Terraform infrastructure.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/goodsmileduck/claude-registry/cloudflare-access-mcp
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent
npx skills add Goodsmileduck/claude-registry --skill cloudflare-access-mcp
Clone the repo
git clone --depth 1 https://github.com/Goodsmileduck/claude-registry

Made for: Claude Code.

Or install cloud-platform-skills, the plugin that ships this one along with the rest of its 7 skills.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for cloudflare-access-mcp

README.md
[![agentmods](https://agentmods.dev/badge/skills/goodsmileduck/claude-registry/cloudflare-access-mcp/github.svg)](https://agentmods.dev/skills/goodsmileduck/claude-registry/cloudflare-access-mcp)
Your own site
<a href="https://agentmods.dev/skills/goodsmileduck/claude-registry/cloudflare-access-mcp"><img src="https://agentmods.dev/badge/skills/goodsmileduck/claude-registry/cloudflare-access-mcp/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for cloudflare-access-mcp

Your own site · 80×15
<a href="https://agentmods.dev/skills/goodsmileduck/claude-registry/cloudflare-access-mcp"><img src="https://agentmods.dev/badge/skills/goodsmileduck/claude-registry/cloudflare-access-mcp.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 164 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,279 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 1 finding. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00164 $0.02279
Opus 5 $0.00082 $0.01140
Sonnet 5 $0.00033 $0.00456
Haiku 4.5 $0.00016 $0.00228

Measured 9d ago against content hash befca38b066f, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-09, from the pricing page.

Security

Grade A, and why

cloudflare-access-mcp scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 9d ago.

The scan reads SKILL.md. This mod also ships 1 executable file (scripts/validate_access_mcp.py), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

- [ ] `curl -s https://<mcp-host>/.well-known/oauth-authorization-server | jq .` returns issuer + authorization_endpoint + token_endpoint + registration_endpoint.
plugins/cloud-platform-skills/skills/cloudflare-access-mcp/SKILL.md · 126 lines

How it starts

The opening of the file, as written. The whole thing — 126 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Cloudflare → MCP server OAuth/SSO

Three real paths, pick by use case.

Decision matrix

Path When Effort Limit
A — AI Controls MCP Portal (REST) You have one or more MCP URLs that already exist (public Workers MCPs, vendor MCPs). You want one portal URL fronting them with OAuth, plus per-tool/per-prompt toggles and Access logging. 1 IdP step + 1 POST per server + 1 dashboard step auth_type is oauth / bearer / unauthenticated only — no Cloudflare Access service-token (dual-header) auth to upstream. Portal composition is dashboard-driven; only server registration is REST today.
B — Self-hosted Access app + Managed OAuth (REST) Per-hostname OAuth on your own origin. No portal hop. Origin can be private (behind cloudflared tunnel) or any public URL. Native MCP-spec OAuth (DCR, /authorize, /token, .well-known/*) at your hostname. ~6 API calls per server (+ tunnel if private) More moving parts. Five known footguns — see Hard rules.
C — Same as B, in Terraform/OpenTofu You already have IaC for Cloudflare. Multi-environment parity matters. PR-reviewed access changes. Heaviest setup, smallest per-server delta Wrong for one-offs. If a project has no Cloudflare IaC yet, pick A or B.

The cf CLI also exposes these endpoints — preview only; the REST recipes here are the stable contract. See the cloudflare-cf-cli skill for cf posture.

When to invoke

  • "Put my MCP server behind Cloudflare", "add OAuth to my MCP server", "expose private MCP server", "MCP portal", "AI Controls MCP", "MCP Managed OAuth", "MCP DCR".
  • Adding another MCP server to an existing setup (Path A: another POST; Path B/C: another API call / tfvars entry).
  • Choosing between paths — present the matrix above.

Inputs to collect first

Input Why Example
Cloudflare account ID + zone ID Account- and zone-scoped resources from dash sidebar
MCP URL(s) Path A: the upstream hostname. Path B/C: the public FQDN you'll publish https://mcp.github.com/mcp (A) or notion-mcp.example.com (B/C)
Path B/C: origin reachability from cloudflared Tunnel ingress service (only if origin is private) http://notion-mcp:3100
IdP + client id + secret location Cloudflare needs an Identity Provider before any path works Google; AWS SM cf/google-oauth
Zero Trust team domain <team>.cloudflareaccess.com; permanent; dashboard mycorp
Allowlist Who can sign in ["example.com"]
Session duration Edge auth cache TTL "24h"

Read the full file on GitHub · 126 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 9d ago First seen · 126 lines · 164 tokens per session scan A befca38b066f

Subscribe to this mod's changes

cloudflare-access-mcp is a skill published in the GitHub repository Goodsmileduck/claude-registry (1 stars, last pushed 1mo ago), licensed MIT. It adds 164 tokens to every session and 2,279 once invoked, about $0.0008 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.