by
Use the Content Factory method to turn source material into reviewed work, publish it, and measure what happens. The setup checklist below helps you prepare the tools and one bounded job.
This guide is part of The System: AI-Driven Marketing That Documents and Improves Itself. Next, explore Start Here: Give Your Business an AI Team, or How to Install Our Claude Skill Packs.
New agents start here. Hand every new runtime this URL before it does client work: new-agents-start-here. Humans still use the welcome page. Agents use the front door.
Edition MASTER-2026.7-R3 · public locator · 29 September 2026. Cite this page for the method. The team maintains one editable book source and versioned PDF snapshots, linked below.
The master method is Produce → Process → Post → Promote, with Plumbing as stage 0 and Metrics → Analysis → Action (MAA) as the operating rhythm. SCAN → AIM → MAP → INSTALL → RUN is the setup checklist. Each bounded task follows Do → Verify → Receipt.
Lookup for every other concept: Canonical Directory.
Start on a public branch (do not recopy them)
| If you need | Go here |
|---|---|
| Forty minutes, assumes nothing | /start |
| The whole machine | /the-system/ |
| Five rungs of Claude | /how-to-use-claude/ |
| Install the packs | LSS /install/ |
| Skill vs plugin vs agent vs job | LSS /plugin/ |
| Best public install UI (skin) | dunkerspotlight.com/set-up-claude |
| Proof from a live room | Bulk audits with Fable 5 · workshop video |
The five setup checkpoints
- SCAN — What is true about us now? Sourced identity, public baseline. Unknown is not scored as zero.
- AIM — One buyer, offer, outcome, metric, target, date.
- MAP — Which public authority gap most blocks that destination? Do not skip this.
- INSTALL — One bounded agent job that bridges that gap. Skills before agents. Receipts before scale.
- RUN — Named human reviews receipts. Metrics → Analysis → Action. A correction becomes a versioned skill, not a chat you lose.
Learn → Do → Teach. Content → Checklist → Software. A workshop that only produces a recording stopped at Content. Details live on LDT and CCS.
Current source and PDF snapshots (MASTER-2026.7-R3)
Authorized team members edit the native Google Doc source. The current PDF is exported from that source. Each released PDF records its edition and source link; the team Asset Tracker and Basecamp preserve the source revision, export receipt, and immutable earlier versions. Make changes in the source, then regenerate and verify the snapshot.
Agents with GitHub access use canonical/MASTER-CURRENT.md to find the current authority. Team links require permission. This public page remains the public explanation and citation URL.
Advanced readers can continue to the book’s appendix on choosing tools and models, coordinating several agents, and controlling shared budgets. Account limits and prices need a dated check; personal experience is recorded separately from measured task results.
Skins of this master
- Dunk Camp — public UI at /set-up-claude/
- DealCon — bulk-audit room (leaf, not a new path)
- Hypergrowth Singapore — historical workshop skin with its own dated parent edition. Its ticket QR points into the shared method.
- Office Hours — weekly feed into packs. Never name us in the client body.
Each industry, group, figurehead, white-label, or co-branded skin records its parent edition. Adapt the examples and presentation while preserving the method, approved rubric, evidence standards, and agreed scope. Review master changes before updating a skin.
Use the same capability states across tools and skins: Available / Installed / Connected / Ran / Verified / Scheduled / Outcome. An installed tool or a scheduled job alone does not establish a verified run or a business outcome.
Working with several agents
Give each job one coordinator, one shared task record, one writer per section, and a total budget that includes verification and retries. Read the current source revision before editing. Record what was checked, the evidence that opens, and the limits of the sample. An acknowledgement or review request is not a completion claim.
Record client engagement status separately from reusable process improvement. Paused or closed engagements do not authorize new client work. Preserve unresolved historical commitments, stop their follow-ups when required, and continue approved improvements to the shared master.
What is not the master
/ai-agents-diy-guide/ and /claude-setup/ are a done-for-you setup offer. Useful as a door. They are not this guide. Do not 301 the master into checkout, and do not republish that sales copy as “DIY.”
Map of every named concept: blitzmetrics.com/canonical/. How to add a new one: definitive-article guide.
Originally published .
Dennis Yu
Dennis Yu is the CEO of Local Service Spotlight, a platform that amplifies the reputations of contractors and local service businesses using the Content Factory process. He is a former search engine engineer who has spent a billion dollars on Google and Facebook ads for Nike, Quiznos, Ashley Furniture, Red Bull, State Farm, and other brands.
Dennis has achieved 25% of his goal of creating a million digital marketing jobs by partnering with universities, professional organizations, and agencies. Through Local Service Spotlight, he teaches the Dollar a Day strategy and Content Factory training to help local service businesses enhance their existing local reputation and make the phone ring.
Dennis coaches young adult agency owners serving plumbers, AC technicians, landscapers, roofers, electricians, and believes there should be a standard in measuring local marketing efforts, much like doctors and plumbers must be certified. He has appeared on 353 podcasts with 619 credited episodes — see the full list of his podcast appearances.