Illustrative Examples — Skill content shown is representative, not from a live build

Examples · What you get

What a skill actually looks like.

Three real shapes. Same standard. Works in ChatGPT, Claude, Gemini & Codex.

ChatGPT Claude Gemini Codex
· Built once · Lives everywhere · Open SKILL.md standard

Three sources. Three skill shapes.

Every skill answers from your pages — with citations.

What you upload 01 / 03
Thinking, Fast and Slow
499 pp · English · Standard quality · Book / Non-fiction
Sample question this skill answers
"What does Kahneman say about anchoring in pricing, and when does it apply?"
thinking-fast-and-slow/ v1.0 · 5 files
SKILL.md          # index + core frameworks
chapters/         # system 1 vs 2, heuristics…
glossary.md       # anchoring, availability…
cheatsheet.md     # decision rules & biases
patterns.md       # dual-process patterns

Your skill answers from YOUR pages — with citations.
Generic AI answers from Wikipedia.

What you upload 02 / 03
Acme Brand Voice
44 pp · English · Standard quality · Brand guide
Sample question this skill answers
"Rewrite this CTA in our voice — what rules apply?"
acme-brand-voice/ v1.0 · 5 files
SKILL.md          # voice pillars + index
chapters/         # tone, register, audience…
glossary.md       # brand vocabulary
cheatsheet.md     # do / don't quick rules
patterns.md       # CTA, headline, copy patterns

Your skill answers from YOUR pages — with citations.
Generic AI writes in default AI voice.

What you upload 03 / 03
SDR Qualification Playbook
62 pp · English · Standard quality · Playbook / SOP
Sample question this skill answers
"Qualify this inbound lead the way we do it."
sdr-qualification-playbook/ v1.0 · 5 files
SKILL.md          # framework + ICP index
chapters/         # MEDDIC, objections…
glossary.md       # ICP, BANT, champion…
cheatsheet.md     # qualification checklist
patterns.md       # discovery call patterns

Your skill answers from YOUR pages — with citations.
Generic AI gives you a generic MEDDIC template.

Examples are illustrative — your skill is built from your source, not these.

Inside every skill

Five files. Each with a reason to exist.

SKILL.md

Index + core frameworks

The AI reads this first. Contains the primary frameworks, use-when triggers, and the navigation map for the rest of the pack.

chapters/

Your knowledge, on demand

One chapter per topic cluster. The AI loads only the chapter that matches the question — focused, not a dump of everything.

glossary.md

Every key term, defined

Prevents the AI from substituting generic definitions for your specific meaning. Your vocabulary, locked in.

cheatsheet.md

Quick rules & decision tables

Fast-access rules the AI reaches for under pressure. Do / don't. When to escalate. The calls that used to be judgment — now consistent.

patterns.md

How you actually work

Recurring structures, templates, moves that characterise your method. The difference between an AI that follows a framework and one that thinks in it.

Most "add your docs" tools toss everything in and hope.
A skill is engineered — every chapter has a reason to exist, and the AI knows exactly where to look. Same SKILL.md standard works in any AI that reads it. Built once. Lives everywhere.

SKILL.md is the open format Anthropic shipped Dec 2025 · OpenAI and Google adopted it the same month

More: who builds skills → · how this compares to DIY →

Build your first skill.

Drop in any book, brand guide, or playbook you own. We build it in 24h. Your AI never starts from scratch again.

Build my first skill

$10 · 7-day trial · Cancel any time · Skills are yours, forever