Meet Bloop: How I Turned My Illustration Style Into a Reusable AI Skill

Every illustration on this blog stars the same character: Bloop, a small blue slime who actually performs the idea I'm writing about. That consistency isn't me redrawing him from scratch each time — it's a skill, in the literal AI sense. One reusable recipe an agent follows so Bloop comes out identical, on-brand, every single time.

Why a mascot, and why a skill

A recurring character does for a feed what a logo does for a brand: people start recognising it. But "draw the same slime again" is exactly the kind of thing that quietly drifts — slightly different blue here, a different vibe there — until the recognition is gone.

So I stopped relying on memory and encoded it. The character, the visual rules, the voice, and the step-by-step workflow all live in one Markdown file. Pick a topic, run the skill, and out comes a finished content package in about two minutes.

What it actually does

Each run hands me three things I can use immediately: a fresh on-brand topic (checked against a log so it never repeats), the caption in my voice, and a ready-to-paste image prompt. The one rule it never breaks: Bloop has to be the mechanism in the scene — becoming the funnel, plugging the leak, filtering the stream — not a smiling decoration in the corner.

Same Bloop, any topic

Here's the proof. Three posts, three completely different subjects — plugin bloat, reviewing AI-written code, and a one-line change that spawns three new bugs. Same slime, same blue, same hand-drawn style every time. That sameness is the skill doing its job.

Grab the skill

It's just a plain SKILL.md in the agent-skill format — no code, no dependencies. Copy it, swap Bloop and the voice for your own character and brand, and you've got a consistent illustration engine of your own. Here's the whole thing:

“One character, one style, infinite topics. That's the whole trick — and now it's a file you can reuse.”