// DEEP DIVES · READER MAP
Choose your way in.
Eleven practical guides, organised by the question you are trying to answer: how to move work forward, which tools to use, how to verify what is true, or which reusable pattern to copy next.
Need the working artifact? Browse all templates →
01 / FROM INTENT TO ARTIFACT
Workflows
How spoken ideas, rough briefs, and open questions become useful work that can be continued.
03 · WORKFLOW
Voice-driven development
Turn spoken intent into reliable instructions an agent can follow.
08 · WORKFLOW
From rough brief to published page
Move a rough idea through structure, visuals, connections, and publication.
11 · SCOPING
How to scope an AI agent task
Turn a loose request into a bounded AI agent task without forcing the agent to guess or drift.
02 / CHOOSE THE RIGHT LEVER
Tools and models
Practical choices about agents, models, settings, and local tools—and the trade-offs behind them.
01 · TOOLS
The four-agent toolbench
Choose the right agent for the job and keep the working context aligned.
02 · SETTINGS
Tuning the knobs: models & settings
Tune model and tool settings for the kind of work in front of you.
05 · LOCAL
Local & uncensored models
Understand when and how local models earn their place in the stack.
03 / TRUST THE EVIDENCE
Verification and context
Methods for giving agents the right context, finding what is true, and checking the result.
06 · METHOD
The verification loop
Use an evidence ladder to turn plausible work into a trustworthy result.
07 · CONTEXT
Context without oversharing
Give an agent enough context to work without publishing what should stay private.
09 · ORIENTATION
Before the AI edits: find the live truth
Trace a live result back to the exact source that should change.
04 / COPY, ADAPT, CONTINUE
Recipes and templates
Reusable structures, packets, and working patterns that can be applied immediately.