AI Codex
The practitioner's guide to deploying AI.
Plain, current guides for rolling out Claude at work and building with it. Which plan to buy, what admins can control, what it really costs, what breaks in month six, and what changed this week. Rechecked against Anthropic’s docs.
New to Claude? Start with the most common mistakes and what’s safe to share.
Where do you want to start?
Rolling out Claude at your company?
Team or Enterprise, what the admin controls actually give you, the usage policy, and getting it through security. The decisions in the order you have to make them.
Using Claude at work?
What is safe to paste in, better prompting, and the workflows that survive past week three — the habits that separate people who get consistent value from Claude from those who don't.
Building with Claude?
From your first API call to production deployment. Or from a product idea to real users — two paths depending on whether you're a developer or a founder.
Going into companies to build it?
Forward Deployed Engineers get hired to build AI systems inside someone else's company. What the role is, Anthropic's new FDE training, the portfolio that gets you interviews, and the three parts of an engagement nobody writes down.
Not sure which one fits you?
Answer 2 questions and we'll point you to the right path.
Worth reading
Worth reading now
If you run Claude Team or Enterprise: Computer use arrived in beta on November 20. It has no sandbox, so set these limits before people turn it on.
Computer use on Claude Team and Enterprise: what admins should set up first
Computer use lets Claude see a person's screen and work in desktop apps. It's in beta on Team and Enterprise, off by default, and has no sandbox. Here's what it can reach, the settings that limit it, and a safe way to roll it out.
5 min read
If you use Claude at work: The question every new user asks first, answered for each plan.
What's safe to share with Claude, and what isn't
The most common reason people hold back from using Claude for real work is not knowing whether pasting something in puts it at risk. Here is the direct answer for each plan as of November 2026, including the training setting most Pro users have never checked.
7 min read
If you build on the API: The new API version rejects temperature, top_p and top_k. Check before you upgrade the SDK.
API version 2026-10-21: temperature is gone, and your SDK upgrade is a major version
On October 21, 2026 Anthropic released the first new Claude API version since 2023-06-01, along with new major versions of every SDK. Requests that send temperature, top_p, or top_k now fail with a 400. API keys move to the Authorization header. Here is what breaks, what to change, and how to upgrade without an outage.
9 min read
If you want to be a forward deployed engineer: Anthropic just published its own FDE learning path. What is in it, and the free version if you are not at a partner.
Anthropic's Forward Deployed Engineer learning path: what's in it and who can take it
Anthropic now has a free, self-paced Claude Forward Deployed Engineer learning path of about 9.5 hours. It is only open to Claude Partner Network members. Here is what it covers, how to get the same material for free if you are not a partner, and how it fits with the exams and the Frontier Academy residency.
6 min read
Free from Anthropic
Anthropic launched Claude Academy. Go take it.
29 free courses plus hundreds of tutorials and use cases — the best product training any AI company has published. We mapped the whole catalog: which six resources matter for your role, in what order, and the operating problems each track leaves you to solve alone.
Or browse by topic
Full glossary →Weekly brief
For people actually using Claude at work.
Each week: one thing Claude can do in your work that most people haven't figured out yet — plus the failure modes to avoid. No tutorials. No hype.
No spam. Unsubscribe anytime.