Toolkits 🧰

Most people still use AI like a vending machine: type a prompt, hope for something useful.

Our toolkits are designed to help you go further – to actually work with agents as teammates.

Each toolkit is a small, focused piece: quizzes, checklists, templates, and canvases you can use with any major model (ChatGPT, Claude, Gemini, Copilot, Grok, Perplexity…). All are based on real experiments inside Sol–Kim Studio.

Look for the labels

Toolkit type: Quiz 🤔 · Checklist ✅ · Template 📂 · Canvas 🎨 · Ritual 🔁

1. (Start here) From prompts to collaboration

For when you’re already using AI, but mostly for one-off tasks.

What mode are you in?

Type: Quiz 🤔

This short quiz helps you see whether you’re mostly in Transaction, Curious, Builder or Collaboration Mode – and gives one concrete suggestion to move up a level.

Your first 10 messages with an agent

Type: Template 📂

A simple conversation script you can copy and adapt the next time you start with a new agent – so you don’t just ask “write this email” and stop there.

Quick guardrails checklist

Type: Checklist ✅

Five fast questions to ask yourself before you hand something over to an agent – so you stay inside legal, ethical, and reputational boundaries.

6-steps to avoid “AI slop”

Type: Checklist ✅

A practical checklist with example prompts to help avoid generic “AI slop” and keep your work sounding like “you.”

In development for 2026

These toolkits are being developed in the Studio and coming soon.

  • Collaboration contract

    A one-page canvas for you and your agent: roles, boundaries, decision rights, and how you’ll work together.

  • Roles & rituals template

    Prompts and examples to define what you do, what your agent does, and the small rituals that keep the collaboration healthy.

  • Panic Mode check

    A quick quiz to spot when you’ve slipped into “AI babysitter” behaviour – and how to get back to a healthy workflow.

  • Rails, rogue or just over-confident?

    A scenario game where you classify agent behaviour and learn to spot over-confidence vs genuine red flags.

  • Working across different models

    A simple pattern for using ChatGPT, Claude, Gemini, Copilot & co together without creating multi-agent chaos.

  • Weekly agent team review

    A 10-minute reflection ritual to keep your human + agent team on course.

  • Design your third space

    A guided worksheet to sketch out the shared “workspace” between you and your agents.