Who this is
We teach it because we do it.
Not the other way round. The Claude Guys exists because we kept building the same things for clients and realised the files were more useful in public than in a private repo.
The two of us
Two engineers, and no one else.
There is no team behind us. You get one of these two on every call, and the one who scoped it is the one who builds it.
Austin Black
AI Solution Architect
Austin has spent his career in information intelligence and B2B enterprise SaaS sales at executive level, on a systems analysis foundation.
As solutions architect he scopes every engagement and decides what it actually is, turning business requirements into architecture that holds up in production.
Scotty Pittsford
AI Development Engineer
Scotty has spent the last 10 years in tech working in data analysis, development, and large language model fine tuning.
He architects backend systems, automation pipelines, and AI infrastructure, turning strategic roadmaps into production-grade solutions that run reliably at scale.
Two names, one team
You will see both. Here is which is which.
The Claude Guys
The teaching half
The library, the newsletter, the community, the workshops. Aimed at individuals — founders, marketers, operators — who want to run their own work on Claude. Everything here is free or cheap on purpose.
Cascadian
The building half
Forward-deployed agent engineering for companies. Audits, fixed-scope projects, monthly retainers. When a workshop turns into “can you just build this for us,” that conversation happens under the Cascadian name and a Cascadian invoice.
The operating policy
Every engagement exits with a reusable file.
A client needed a brand name screened for trademark conflicts and domain availability. We built it, it worked, and on the way out we stripped the client specifics and kept the method. That became four skills anyone can install today.
That is the whole engine. The library grows as a byproduct of paid work, which means it grows in the direction of things people actually pay to have solved — and it is a moat a content business structurally cannot copy, because they have no engagements to genericize.
Every asset in the library was used on a paying client engagement.
What we will and won’t do
The rules we hold ourselves to, in public.
Every number is one we can show you the file for
If a count on this site looks round, assume we rounded down. We publish the real inventory and update it when it changes.
No credential we don't hold
We are pursuing Anthropic's certifications through the Claude Partner Network. Until a badge exists and is publicly verifiable, it does not appear on this site.
Attribution stays intact
Curated open-source keeps its upstream licence and its author's name. We do not rebrand other people's work, and a pre-publish scan blocks licence-flagged material from shipping.
Our scarcity is real or absent
We do not run countdowns that reset, or seat counts that never fall. If we ever publish a limit it will be one you can check — against the live number, or against this page in the Wayback Machine.
No client named without permission
We will not print a client's name, logo, or rate to make ourselves look bigger. Case studies happen when the client says yes, in writing.
We publish what didn't work
The failures are more useful than the wins and nobody selling a course can afford to write them down.
The library is the fastest way to find out if we’re any good.
The assessment
How good are you at Claude, actually?
Eight questions. You get the level you actually test at, the one move that gets you to the next one, how many hours a year the job you name is costing you, and three skill files cut to your title, company and industry.
Take the testUnder two minutes · no account · change any answer
- 01The Beginner
- 02The Chat Master
- 03The Operator
- 04The Automator
- 05The Agent Conductor
Section surveyed 5,013 knowledge workers: 54% rate themselves AI proficient, about 10% test as proficient.