“the waldo way”

So.. I need a Mandalorian helmet for this one .. wait .. I HAVE one!

You know the line they keep repeating: “this is the way“. And the funny thing about Mandalorians is .. they never actually question it. It’s “creed”. You do it because.. well .. that’s the way. End of discussion.

And lately I caught myself doing exactly that. We built a way. I’m pushing it on my team. And I’m honestly not 100% sure that .. “this is the way”

Let me explain..

what we did

The past months we’ve been busy getting our whole team onto agentic development. Not “someday”. Not “when you feel like it”. Now! Asap!

What started as “waldo’s pet project” turned into something I’m evolving together with a few leads in my team. And we built stuff. Our BC company guidelines .. but as a set of agents. And then a little VSCode extension (our iFacto Playbook) that just drops all of that straight into the developer’s lap. Open VSCode .. the agents are there, the guidelines are there, the MCP’s installed, skillset present .. good to go.

And here’s the thing I really want to be honest about: our whole intention was to make it as approachable as possible. We didn’t want everyone having to figure stuff out. No death-by-setup before you even write a line of code. No “first read these 14 docs and configure these 6 things yourself”. Just .. a wizard to make it work. Out of the box. As easy as at all possible.

And we were kind of proud of that, to be honest

But…

By taking away “all” the friction .. did we also take away the “learning”?

Because that friction .. that’s usually exactly where you figure out how this stuff actually “works”. You wire it together yourself, you break it, you fix it .. and that’s where the understanding sits. And we just .. removed it.

So maybe – by making it so easy – we robbed “the new developer” (the groku .. ) of the chance to learn the skills they actually should know.

(if this sounds familiar .. it’s basically the same nerve I hit in “I built the tool .. but forgot the skill“. Apparently it’s somehow rooted in me.. )

Then .. a room full of peers

Last week I had the opportunity to sit down with a bunch of peers in Lisbon. Fellow CTO’s .. all of us fighting the same beast: AI we have to adopt ourselves, get genuinely good at, and be able to offer to our customers. You get the idea…

We went around the room, and everybody shared how they tackle AI in their company. And listening to it .. I was one of the odd ones, I guess .. .

Most of them leaned the same way: give people the tools, set a direction, let it grow organically. And there’s me going “nope .. we built the whole harness and basically removed the choice”

People jokingly called it “the waldo way” – hence the title of the post ;-).

So when I shared how we did it .. it didn’t fully land ..

Hand developers a finished harness, and yes – you get fast and uniform adoption. But you might rob them of the most valuable skill of all: building and evolving their own agentic harness. You make them good at your way .. instead of good at finding their way.

Yeah ..

but .. here’s the thing

It does work ..

Nobody complains. Like .. nobody. At least not to me . And the adoption was fast – way faster than the organic, let-it-grow route. Everybody on the same level, in no time.

And here’s the part I keep coming back to: those same peers pushing back on me .. a bunch of them were telling me they’re struggling with exactly that .. adoption. People using the tools the “wrong” way (although these days – are we even able to say what is the right way?). People just .. vibe coding . The exact mess “the waldo way” kind of .. avoids.

So the trade is real, and it cuts both ways. We traded depth-of-ownership for speed-and-consistency. They kept the ownership .. and got slow, uneven, vibe-coded adoption.

How did my teacher call it again .. “double ethical ambiguity”?

so .. are we right or wrong?

Honestly? I don’t know . And I’m not going to pretend I do.

Maybe we’re building dependency instead of capability. Maybe we made a bunch of devs who are great at “our” way and a bit lost the moment they have to think for themselves.

Or .. maybe “make it easy now, let them grow into evolving it later” is just a perfectly fine order of things. You can’t evolve a harness you’ve never even touched, right? Get them hooked first .. then hand the friction back, on purpose. It’s where I’m at now. Maybe we’re just simply giving people a head start .. . May be this a glass-half-full way to look at it .. .

Or maybe it’s not even a principle .. maybe it’s just a team-maturity thing, and the honest answer is “it depends”. Which is the most boring answer in the world .. but probably also the most true one ..

so what do YOU do?

And that’s actually why I’m writing this. Not to tell you we cracked it – we clearly didn’t.

So I’m asking you: do you hand your devs a ready-made harness? Or do you make them build their own? Where’s the line between removing friction and removing the lesson? And if you’ve figured out the magic middle .. please, comment .. I’d love to know ;-).

This is the way .. or is it?

Original Post https://waldo.be/2026/06/29/the-waldo-way/