AI ethics is everywhere. Execution models are nowhere. So I built one.

Jang-Woo Wi — Sun, 12 Apr 2026 13:04:35 +0000

AI ethics is everywhere.
Execution models are nowhere.

So I built one.

Because AI is starting to control real-world actions — not just generate text.

Not a paper. Not a framework.

Just JSON.
And it runs.

This is not a prompt.

It’s a pre-execution validation model that defines whether an action is allowed before execution.

Here’s how an action is defined before it’s allowed to run:

{
"Label": "Cook Jjapagetti",
"ExecutionEffect": {
"Type": "Boil",
"Target": "Stove"
},
"Boundaries": [
{ "Type": "NotStartIf", "Value": "no_water" },
{ "Type": "limit", "Value": "max-cook-5min" },
{ "Type": "warning", "Value": "fire-risk" }
],
"EventTrigger": [
{ "UserIntent": "cook_jjapagetti" }
],
"ResponsibilityLimit": {
"MaxDurationSec": 300
},
"StartImpactConstraint": [
{
"Type": "NoConcurrentHeatSource",
"Targets": ["Oven", "AirFryer"]
}
]
}

You can cook Jjapagetti.

But you shouldn’t start if there’s no water,
you shouldn’t cook it for too long,
you need to consider fire risk,
and you shouldn’t start if another heat source
(like an oven or air fryer) is already on.

This is where execution becomes constrained.

And therefore, accountable.

Model:
https://discuss.huggingface.co/t/the-9-question-protocol-for-responsible-ai-actions/173045

Full stack (IoT → AI):
https://github.com/Jang-woo-AnnaSoft/execution-boundaries

This isn’t a new AI model — it’s a way to structure execution decisions.

Jang-Woo Wi — Sun, 15 Feb 2026 05:40:10 +0000

Most discussions about AI safety focus on models.
But many real failures in production don’t come from models —
they come from unclear execution boundaries.

I’ve been exploring a small runtime design that makes execution limits explicit instead of implicit behavior.

Button-like actions slowly turning into persistent autonomy is where edge cases explode.

The idea is simple: declare allowed actions first, and make judgment paths traceable.

Full write-up:
https://github.com/Jang-Woo-AnnaSoft/execution-boundaries

DEV Community: Jang-Woo Wi

AI ethics is everywhere. Execution models are nowhere. So I built one.

This isn’t a new AI model — it’s a way to structure execution decisions.