Command note: An undo that overwrites a person's edit is worse than no undo. A rollback is only safe if it knows which values are still yours. Anything a human changed after the build belongs to them now. We shipped this in the current beta of the client-account build: the undo puts a value back only while it still holds exactly what the build last wrote. A value someone edited since is left alone and listed as changed since the build. Where this applies: • Undoing a bulk build or import • Reverting a template or snapshot push • Any cleanup run you do after handoff Do it now: 1) Record the before-value of every field you change — the exact value you saw, at the moment you approved the change. 2) Record what you wrote — so you can tell your change from a person's later edit. 3) At undo time, compare the current value to what you wrote — a match gets restored, a mismatch gets left. 4) List every skipped item — tell the owner what you left and why. A clean undo restores your work. It never erases theirs.

Command note: A pipeline stage no deal ever reaches is a promise your process isn't keeping. Stages are instructions. If nothing ever lands in one, either the trigger that moves deals there is broken or the stage describes a step nobody actually does. What a dead stage usually means: • The automation that should move deals in was never connected • The stage was copied from a template and never fit how you sell • Deals skip it because the real work happens off-pipeline Try this prompt in Claude with the GHL Command MCP connected: "List every pipeline in this sub-account. For each stage, count the open opportunities in it and the date of the most recent one moved in. Flag any stage with zero deals, or none moved in during the last 60 days. Read-only, don't change anything." Do it now: 1) Run the prompt on one sub-account — start with the pipeline that matters most 2) Mark each flagged stage: broken trigger, wrong stage, or dead step 3) Fix the trigger or delete the stage by hand in GHL, one change at a time 4) Re-run next month and compare The prompt only reads. Every fix stays a human decision. A stage nobody reaches isn't a stage. It's a wish.

Command note: If the account only makes sense to the person who built it, it isn't finished. Every build has context that lives in your head: why a stage exists, which workflow is paused on purpose, what the client asked you never to touch. The day you step away, that context walks out with you. What a handoff note covers: • What is live, and what is deliberately switched off • Where the odd decisions are, and why they were made • What the client has asked never to change • Who to ask, and what to check first when something breaks Do it now, in about 20 minutes: 1) Pick one account you could not hand to someone else today 2) Write four short headings: Live, Paused on purpose, Do not touch, First checks 3) Fill each with plain sentences, no jargon only you understand 4) Hand it to someone who has never seen the account and ask them to explain it back 5) Fix every spot where they stumble, then save it where the whole team can find it The test is simple: could a stranger run this account tomorrow using only the note? A build is only as good as the note that lets someone else keep it running.

Command note: A form with zero submissions is not a form. It is a dead end with your logo on it. The forms you forgot are the ones quietly failing. Old embeds, retired offers, a page that got rebuilt and lost its form. Nobody complains because nobody ever arrives. Prompt to run today: "List every form in this sub-account with its submission count for the last 90 days. Flag any with zero. For each flagged form, tell me which workflow and which page it is attached to." What the answer tells you: • Zero submissions, still embedded on a live page: it is broken or buried. Fix the page. • Zero submissions, not embedded anywhere: it is orphaned. Archive it. • Submissions but no workflow attached: leads are landing and nothing follows up. • Two forms collecting the same thing: pick one, retire the other. Do it now: 1) Run the prompt on one active sub-account. 2) Submit each flagged form yourself with a test contact. 3) Confirm the contact lands, gets tagged, and triggers the follow-up. 4) Archive what is orphaned and write down what you kept and why. A quiet form is not a calm form. Check it before you trust it.

Command note: Deleting is easy. Finding out what depended on it is the expensive part. Lesson learned: we once removed a tag we thought was dead. Two workflows still keyed off it, and both quietly stopped firing. Nothing errored. We noticed a week later when follow-ups went missing. The rule: nothing gets deleted until you know what points at it. What quietly depends on other things: • Tags that trigger or filter workflows • Custom fields used in templates and merge fields • Pipeline stages referenced by automations • Trigger links and forms that feed a workflow • Calendars tied to booking confirmations Do it now: 1) List what you plan to delete — one line each 2) Search every workflow, template, and smart list for each name 3) Write down every hit and who owns it 4) Repoint or retire those dependents first 5) Archive or rename before you delete, then wait a week 6) Delete only when nothing has complained A deleted item never sends an error. It just stops doing its job.