Secure MCP Servers in 30 Seconds (AuthKit + OAuth) @WorkOS

Uploaded March 2026 | Updated September 2026, 1 day ago

How do you secure an MCP server?

In this short demo, we take a FastMCP server and secure it using WorkOS AuthKit.

Steps covered:
• Add a small auth block to your FastMCP server
• Enable Dynamic Client Registration in the WorkOS dashboard
• Configure an OAuth2 redirect URI
• Add your Client ID and AuthKit URL
• Restart the server and reconnect the client

Now your MCP server is secured with the same auth platform used by OpenAI, Anthropic, and Cursor.

WorkOS AuthKit includes:
• OAuth and enterprise authentication
• RBAC and FGA for authorization
• Vault for encrypted org-level storage
• Radar for bot and fraud protection

Learn more about securing MCP servers with WorkOS.

Secure MCP Servers in 30 Seconds (AuthKit + OAuth)

Agents Will Choose Your Entire Tech Stack

How AI Agents Are Replacing Configuration — David Crawshaw, exe.dev | Agent Night

Enterprise Authorization Patterns for AI Agents — Garrett Galow, WorkOS | MCP Night

Why Agents Break Your Access Control — Umaimah Khan, Opal Security | Enterprise Ready Conf 2025

Can Your AI Survive Crashes? Fault-Tolerant Gen Code — Maxim Fateev, Temporal | HumanX 2026

AI Is Rewriting Healthcares Hidden Layer — Zack Kanter, Stedi | re:Invent 2025

Agent Night: Is AGI Already Here? Jaya Gupta, Flo Crivello & swyx on AI Agents

Lower Your Incident Cost to Ship Faster — Chris Evans, Incident.io | re:Invent 2025

$10x the Tokens, Only 2x the Output — Nicholas Arcolano, Jellyfish | AI Engineer Worlds Fair 2026 Spending 10x more on tokens doesnt make an engineering team 10x more productive. Jellyfishs data, drawn from hundreds of thousands of developers, shows the top 10% are about twice as productive as the median while burning 10x the tokens to get there. Nicholas Arcolano, Head of Research at Jellyfish, sat down with WorkOS at the AI Engineer Worlds Fair 2026 to explain what the token-maxing era actually bought, and where it hit a wall. We get into: • Why tokens are rocket fuel and token maxing is rocket science, and the diminishing returns as spend climbs • The distinct regimes of token use, from fancy autocomplete to 500-million-tokens-a-week multi-agent workflows • Why most engineers still spend a fraction of their salary in tokens, and why finance hasnt caught up to treating compute as capacity instead of a dev-tools line item • Why the real bottleneck has moved off coding onto product, review, and go-to-market: teams cant even absorb 2x, let alone 10x • What a genuinely rebuilt product development lifecycle takes, and why most orgs only move that fast in crisis mode • Arcolanos advice for anyone trying to modernize a resistant organization: show, dont tell A grounded, data-backed look at what AI is really doing to engineering capacity, from someone measuring it across roughly 1,000 companies. To learn more about Jellyfish, visit https://jellyfish.co/. To learn more about WorkOS, visit https://workos.com.$

Dwarkesh Patel | Dwarkesh Unplugged presented by WorkOS