The operating loop
A repeatable brief-to-decision loop that makes useful AI work easier to improve and reuse.
Free resource / practical AI
H² Workflow Starter Pack
A concise field guide for turning scattered AI experiments into one useful workflow — with clearer inputs, a sensible human check and a result you can actually measure.
Get the free pack ↗Free via H² Signal. One concise weekly note; unsubscribe anytime.The pack is for people who already know AI can help, but want a calmer way to choose where it belongs in their day-to-day work.
A repeatable brief-to-decision loop that makes useful AI work easier to improve and reuse.
A clear structure for giving AI the context, constraints and output shape a task actually needs.
A practical way to turn scattered sources into a short brief with evidence, trade-offs and a next action.
A simple go/no-go check for deciding whether an automation saves enough time to be worth maintaining.
A 20-minute review for keeping the AI patterns that work and stopping the ones that do not.
Choose work that happens often enough to be worth improving.
Define the useful output and the points where a person needs to check it.
Run one small test, compare the result and keep only what reduces friction.
Copy a block into your next AI session, replace the brackets and keep a person responsible for the final decision.
Use this before opening a chat or automation builder.
Task: [what needs to be done] Audience: [who will use the result] Context: [facts, files or links] Constraints: [time, format, tone, exclusions] Useful output: [what good looks like]
Make the human review explicit instead of assuming the first draft is ready.
Review the draft for: - unsupported claims or missing evidence - unclear assumptions and open questions - tone, format and audience fit - actions that need human approval Return: issues, fixes and a ready/not-ready decision.
Record one small improvement so the workflow gets better next time.
Weekly workflow review: What repeated task did we test? What saved time or improved quality? Where did the output need correction? What one change will we keep next week? Owner: [name] Review date: [date]
Get the Starter Pack, then use it on one task that matters this week.
Need a deeper operating layer? Explore the Operations Pack ↗Send me the free pack ↗