Playbooks
Practical ways of working with AI.
Repeatable workflows, patterns and useful methods - with general AI conceptualisation and practical application.
What Better TV Recommendations Taught Me About Giving AI Context
A simple example of how interactive context gathering can turn a generic AI answer into something genuinely useful.
Read playbook
Using Different AI Models to Critique and Evaluate Each Other
A practical way to use different model families as fresh pairs of eyes for important reasoning, planning and system design.
Read playbook
Building a Skill Architect to Design Better AI Skills
I built a Skill Architect to improve how I create reusable AI skills, using deep research, Codex, Fable to critique and real-world testing. It is still an experiment, but the early signs are promising.
Read playbook
Building an AI-Animated Website Hero From Separate Assets
A practical experiment in using image generation, AI video, editing tools and Codex to build a more ambitious website hero from separate components.
Read playbook
Before the Coding Agent: Build the Project Context First
A practical workflow for turning messy planning conversations into reliable project context before coding begins.
Read playbook
Don’t Fix the Output — Fix What’s Underneath
How to recognise when repeated AI mistakes are symptoms of a deeper problem in the data, rules or source of truth underneath.
Read playbook
AI Skills: Stop Teaching the Agent the Same Job Twice
How AI skills package repeatable ways of working so you stop rebuilding the same instructions every time.
Read playbook
Iteration Discipline: Give AI Enough to See the Destination, Not Enough to Get Lost
How to give coding agents enough context to see the goal while keeping each stage small enough to review and steer.
Read playbook
Why .md Files Matter So Much in AI
How simple Markdown files become a persistent control layer for AI agents—and how to keep that guidance useful.
Read playbook