Thank you Joel. This is an easy to do exercise that brings great leverage to your operation. Between this and your second brain, I am heading to Aruba. 😂
Thank you Joel. I have been doing this on a one off basis but creating repeatable skills is sooo powerful. Ppt Outputs are much better than gamma as well in terms of visualization!
40 hours back is basically getting an extra week every month. The key is treating Claude as a thought partner, not just a task executor. When you structure requests as strategic problems instead of simple commands, the output quality jumps exponentially.
100%. If you treat Claude like a junior task-doer, you’ll get junior outputs. Treat it like a thought partner: give context, constraints, trade-offs, and ask it to challenge your plan.
Example: don’t say write an email say here’s the goal + audience + risks… draft 3 options, predict objections, and recommend the safest one. That’s when 40 hours back becomes real and the next wave is turning these strategic workflows into repeatable Skills/playbooks so good thinking scales, not just typing.
Really cool how Claude Skills work, and love your breakdown Joel. I am trying to also integrate specific formats with my content calendar, so that I can potentially create different kinds of posts. Hoping that will further streamline things for me :)
One of the things I do is create a new brand skill with the brand guide of each business I have, so it knows exactly what to pull depending on the business
This is a great example of AI being used as infrastructure rather than a novelty feature, and what stood out for me is the difference between tools that simply generate output and systems that actually fit how you already work. Most people I see struggle not because the tech is weak, but because they try to bolt AI onto broken workflows and then blame the tool when it feels clunky. When the workflow is designed first and AI is there to carry the heavy lifting, the quality usually goes up rather than down. The real gain here is not speed alone, it is consistency without the mental tax. Where do you think people most underestimate the setup work needed to make this genuinely useful?
This is exactly the shift from AI demos to AI operations: packaging your real SOPs into Skills so proposals, contracts, and decks become repeatable workflows not fresh reinvention every time.
The future moat isn’t prompts, it’s a versioned skill library + guardrails (data boundaries, approvals, audit trail) so teams can scale output without scaling chaos. That’s how 40 hours saved becomes believable because it’s system change, not motivation.
“Projects are context (what Claude knows), Skills are capabilities (what Claude does)”
that alone is a great take away to really clear up a lot of the confusion i see around claude! great post :)
it was a confusion I had!
Great job on this, Joel! I definitely got a few ideas I hadn't considered before. Love what Claude is doing!
Thank you Joel. This is an easy to do exercise that brings great leverage to your operation. Between this and your second brain, I am heading to Aruba. 😂
Haha! Yes that’s the plan!!
Thank you Joel. I have been doing this on a one off basis but creating repeatable skills is sooo powerful. Ppt Outputs are much better than gamma as well in terms of visualization!
I was a heavy gamma user and yes, this is better :)
Genuinely useful. Thank you.
I’m glad, Emmy!
Awesome! Claude is a game changer. Learning to use it well makes all the difference.
It really is! I’ve switched fully from Chat gpt around a yr ago
40 hours back is basically getting an extra week every month. The key is treating Claude as a thought partner, not just a task executor. When you structure requests as strategic problems instead of simple commands, the output quality jumps exponentially.
Anfernee we need to do a live sometime! I think it’d be a fun chat
But I’m still a Notion AI guy. Haha!
Haha I guess we all have all have areas to improve ;)
🤣🤣
Sure 💪🏻
100%. If you treat Claude like a junior task-doer, you’ll get junior outputs. Treat it like a thought partner: give context, constraints, trade-offs, and ask it to challenge your plan.
Example: don’t say write an email say here’s the goal + audience + risks… draft 3 options, predict objections, and recommend the safest one. That’s when 40 hours back becomes real and the next wave is turning these strategic workflows into repeatable Skills/playbooks so good thinking scales, not just typing.
Great example!
This is gold!
Thanks Luis! Glad to help you customize any of this, reach out if you have questions
Love that Joel, I'll test some of these myself
Ask if you have any questions!
will do for sure
This is quite useful, Joel. Thanks for sharing in such great detail.
Glad it helps! That’s my goal
Really cool how Claude Skills work, and love your breakdown Joel. I am trying to also integrate specific formats with my content calendar, so that I can potentially create different kinds of posts. Hoping that will further streamline things for me :)
One of the things I do is create a new brand skill with the brand guide of each business I have, so it knows exactly what to pull depending on the business
That’s a good idea!
Great breakdown, thanks for sharing!
Thanks @Benjamin Hies
Repeatable workflows usually create the biggest productivity gains.
So true!
This is a great example of AI being used as infrastructure rather than a novelty feature, and what stood out for me is the difference between tools that simply generate output and systems that actually fit how you already work. Most people I see struggle not because the tech is weak, but because they try to bolt AI onto broken workflows and then blame the tool when it feels clunky. When the workflow is designed first and AI is there to carry the heavy lifting, the quality usually goes up rather than down. The real gain here is not speed alone, it is consistency without the mental tax. Where do you think people most underestimate the setup work needed to make this genuinely useful?
This is exactly the shift from AI demos to AI operations: packaging your real SOPs into Skills so proposals, contracts, and decks become repeatable workflows not fresh reinvention every time.
The future moat isn’t prompts, it’s a versioned skill library + guardrails (data boundaries, approvals, audit trail) so teams can scale output without scaling chaos. That’s how 40 hours saved becomes believable because it’s system change, not motivation.
What the heck Joel, amazing breakdown here! And perfect and easy to understand explanation of how to install/upload the skill!
And the examples are amazing too, killed it with this post.