by /
Everybody who installs an AI skill file for the first time asks the same question a week later: why does it feel like I still have to do everything?
The answer is that they installed a function and expected a person. Those are different things, and the gap between them is the whole reason most people give up on agents about ten days in. This is the framework we teach for closing it, the diagrams we use to teach it, and the actual schedule running on our side of the table so you can check that we do it too.
The distinction
A skill is a written recipe. Ours are markdown files, usually 80 to 150 lines, telling any capable model how to do one job well — find every good thing anyone ever said about you, or score a page against a 30-point authority scale. A skill is stateless. It is identical for everyone who downloads it. It runs when you ask, and it forgets you the moment it finishes.
An agent is that same file plus four things: a memory of your work, a time it wakes up, keys to the things it may touch, and a written record of what it did. Those four additions are not features. They are the difference between owning a tool and employing somebody.
Related
- The System — how skills, definitive articles, meta articles and agents fit together as one loop.
- The Skill Pack directory — the actual skill files, kept current daily by the propagation agent described above.
- The meta article prompt — the “written record” requirement, as a standard your agents can follow.
- How female experts use AI agents to build authority — the live workshop where this framework was first taught end to end.
The one you feel first is persistence. An agent that remembers last week stops reporting facts and starts reporting changes — and changes are the only thing you can actually act on.
Related
- The System — how skills, definitive articles, meta articles and agents fit together as one loop.
- The Skill Pack directory — the actual skill files, kept current daily by the propagation agent described above.
- The meta article prompt — the “written record” requirement, as a standard your agents can follow.
- How female experts use AI agents to build authority — the live workshop where this framework was first taught end to end.
“You have nineteen mentions” is a report. “You went from fourteen to nineteen, and here is the new one” is a decision. Only the second one is worth a Monday morning, and only an agent with memory can produce it.
Five things turn a function into a person
| What you add | What it buys you | Why it matters |
|---|---|---|
| Persistence | It remembers | It reads its own notes before it starts, so it can report the delta instead of the raw state. |
| A schedule | It shows up | You stop being the trigger. Any task that only happens when you remember it is a task that dies the week you travel. |
| Handoff | It has colleagues | One agent’s output is the next agent’s input. This is the step that turns a pile of tools into leverage. |
| Permissions | It has keys, and limits | You decide what it may touch. Anything that leaves the building — email, posts, money — is drafted, never sent. |
| A record | It can be audited | Every run writes down what it did. That is how you trust it, correct it, and how it improves instead of repeating itself. |
Then you stop managing agents and start managing departments
Seventeen agents is too many things to hold in your head. Six departments is not. So we arrange them the way a company is arranged, for exactly the reason a company is arranged that way: it is the only shape where the person at the top reads one page instead of seventeen.
Related
- The System — how skills, definitive articles, meta articles and agents fit together as one loop.
- The Skill Pack directory — the actual skill files, kept current daily by the propagation agent described above.
- The meta article prompt — the “written record” requirement, as a standard your agents can follow.
- How female experts use AI agents to build authority — the live workshop where this framework was first taught end to end.
The handoffs are the product
Look at the rail under the columns. The reputation gap tells the researcher what to hunt for. The proof becomes page copy. The hub feeds the spokes. The winners get the dollar a day. The spend reports its numbers, and the loop closes. Any one of those agents on its own is a hire. The arrows are the company.
The shared brain is the part people skip
Seventeen agents each keeping private notes is seventeen freelancers who have never met. One folder that all of them read before they start and write to when they finish is a team. It is also the reason agents get better the longer you keep them — every run adds to the same memory. Two months in, they know things about the business that no fresh tool could be briefed on in an afternoon.
Delete that folder and you have seventeen strangers again. This is the single most important architectural decision in the whole setup, and it costs nothing: keep one project folder, named after the person or the business, and make every agent read from it.
Related
- The System — how skills, definitive articles, meta articles and agents fit together as one loop.
- The Skill Pack directory — the actual skill files, kept current daily by the propagation agent described above.
- The meta article prompt — the “written record” requirement, as a standard your agents can follow.
- How female experts use AI agents to build authority — the live workshop where this framework was first taught end to end.
Our own board: 38 jobs, this week
We would not teach this if we did not run it. Here is our live schedule, sorted into the same departments. Client names are omitted on purpose; the shape is the point.
| Department | Jobs | Cadence | What runs there |
|---|---|---|---|
| Front desk | 5 | Every day | Order forms, new-member registrations, website requests and yesterday’s call recordings get read, sorted and turned into work before anyone opens a laptop. |
| Research | 8 | Weekly & monthly | One agent per client re-pulls rankings, AI-answer citations and audit scores, then writes what moved and why. |
| Build | 4 | Weekly & monthly | Wikidata eligibility, Grokipedia readiness, structured-data sweeps, entity refreshes — the plumbing behind Knowledge Panels. |
| Publishing | 3 | Daily & weekly | Article waves, video cutting, and an editorial interlinking pass that links our sites only where it genuinely helps a reader. |
| Growth | 2 | Weekly | Dollar-a-day performance, and a scan for new opportunities across the network. |
| Quality & money | 14 | Every 3 hours → monthly | Uptime probes, security drift, plugin-licence QA, a Friday audit that re-checks last week’s own decisions, and the job that rewrites the skill files themselves. |
| Reporting | 2 | Friday & Sunday | The weekly MAA into every client dashboard, and the personal-brand build update. |
Notice where the weight sits. The largest department is not the one that makes things — it is the one that checks things. Fourteen of thirty-eight jobs exist only to catch problems and improve the other jobs. One of them rewrites the skill files every single day, which means the files people downloaded from us last month are not the files that are running today. They inherit the improvements automatically, provided they set one job: the Monday refresh.
Related
- The System — how skills, definitive articles, meta articles and agents fit together as one loop.
- The Skill Pack directory — the actual skill files, kept current daily by the propagation agent described above.
- The meta article prompt — the “written record” requirement, as a standard your agents can follow.
- How female experts use AI agents to build authority — the live workshop where this framework was first taught end to end.
How to actually set one up
This is the part that surprises people. You do not open a settings panel and you do not write a cron expression. You say it, in a sentence, and the agent creates the schedule itself while it carries on with whatever else you asked for.
The first one to set, for anybody, is maintenance:
“Every Monday at 8am, go and get the newest versions of my skill files, check which ones changed, and update mine. Set that up as a scheduled task so I never have to remember it. Tell me when it’s done and what you set up.”
Related
- The System — how skills, definitive articles, meta articles and agents fit together as one loop.
- The Skill Pack directory — the actual skill files, kept current daily by the propagation agent described above.
- The meta article prompt — the “written record” requirement, as a standard your agents can follow.
- How female experts use AI agents to build authority — the live workshop where this framework was first taught end to end.
The second is your first real employee — for most people, proof research, because almost everyone is sitting on more credibility than they have collected:
“Every Monday at 9am, find anything new anyone said about me in the last week — reviews, mentions, tags, podcast appearances, comments. Add what is new to my proof library, and write me a short note with only what changed since last Monday. Do not send anything to anyone.”
Related
- The System — how skills, definitive articles, meta articles and agents fit together as one loop.
- The Skill Pack directory — the actual skill files, kept current daily by the propagation agent described above.
- The meta article prompt — the “written record” requirement, as a standard your agents can follow.
- How female experts use AI agents to build authority — the live workshop where this framework was first taught end to end.
That last sentence is the house rule, and it should be in every scheduled task you ever write. Reading, researching, scoring and drafting happen freely. Anything that leaves the building gets drafted and waits for a human. Money never moves at all.
The failure you will hit in week one
You will hand an agent three hours of work, come back, and find it stopped after ten minutes to ask whether you wanted A or B. Models are trained to be careful, and careful looks like paralysis when nobody is watching. The fix is one line, stated up front:
“Do not stop for small decisions. Choose sensibly, keep working, and list every choice you made at the end.”
Related
- The System — how skills, definitive articles, meta articles and agents fit together as one loop.
- The Skill Pack directory — the actual skill files, kept current daily by the propagation agent described above.
- The meta article prompt — the “written record” requirement, as a standard your agents can follow.
- How female experts use AI agents to build authority — the live workshop where this framework was first taught end to end.
We ship that instruction as its own file, called Boil the Ocean, because it comes up so often it deserved a name.
The test
Here is how to know whether you have agents or just files. Go on holiday for a week and do not open your laptop. If work happened while you were gone — if there is a report waiting, a list of drafts to approve, a number that moved — you have a team. If nothing happened, you have a very good library, and libraries do not scale a business.
You already hired them. The only thing left is to tell them when to come to work.
Related
- The System — how skills, definitive articles, meta articles and agents fit together as one loop.
- The Skill Pack directory — the actual skill files, kept current daily by the propagation agent described above.
- The meta article prompt — the “written record” requirement, as a standard your agents can follow.
- How female experts use AI agents to build authority — the live workshop where this framework was first taught end to end.
Related
- The System — how skills, definitive articles, meta articles and agents fit together as one loop.
- The Skill Pack directory — the actual skill files, kept current daily by the propagation agent described above.
- The meta article prompt — the “written record” requirement, as a standard your agents can follow.
- How female experts use AI agents to build authority — the live workshop where this framework was first taught end to end.
The operations half of this argument: Persistent Agents — why a skill pack does no work. This page is the anatomy: what you wrap around a skill file to make it an employee. That page is how the job is actually built and how you find out it is lying to you — the schedule, the QA cycle, the working files, the six-question test, and two of our own failures where every report stayed green.
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.