I turned my consulting playbook into skills the client runs

The dirty secret of consulting is how repetitive the expert part is.

I run an AI-native Attio practice. For the first stretch of it, every engagement had me doing the same handful of things by hand. Open the pipeline, find the deals that had gone quiet, and nudge the owner. Read last week's call recordings and pull out the lines worth turning into content. Prep a one-pager before each sales call so the rep walked in knowing the account. Clean the records nobody had touched. Listen to a rep's call and tell them what to do differently next time.

None of that was hard for me. It was the judgment I had built up over years, applied again. But it only ran when I ran it. The second an engagement ended, every one of those loops stopped, and the workspace started drifting back to the state it was in before I showed up. I was selling a result that depended on my presence to exist. That is the trap of consulting, and for a long time I did not see it as a trap. I saw it as the job.

This is what I did about it, and what changed when I did.

The reframe that broke it open

A skill is a piece of judgment written down precisely enough that something else can run it without you.

In Claude Code, a skill is a folder of markdown. It says what to read, what to ignore, what to decide, what to write back, and in what order. I ship them with no server and no API key of mine: they run on the client's own Claude Code subscription, on a schedule, against their own CRM. That post is about where the agent runs. This one is about what is inside it, because the inside is the thing I used to sell by the hour.

Once I saw a skill that way, the question stopped being "what AI feature should I build" and became "which part of what I do by hand is actually a repeatable decision." Most of it was. The parts that were not turned out to be the parts worth keeping for myself, which I will get to.

What was in the playbook

I went through the things I did on every engagement and asked, for each one, whether it was judgment I applied the same way every time. The ones that were became skills.

• Chasing stale deals. Every week I scanned the pipeline for deals that had gone quiet past the point where quiet means dead, and flagged them to the owner with the context of why they mattered. That is a rule plus a judgment call about what "stale" means for this team. It is now a pipeline-hygiene skill that runs every Monday morning.
• Pulling content out of calls. After calls, I read transcripts for the lines worth writing about: the exact pain a buyer described, an objection, a phrase they used that I would never have invented. That became a content-mining skill that reads the week's recordings, writes the seeds into a Notion backlog, and sends a short digest on Friday.
• Keeping the team focused on the right deals. The judgment of which open deals deserve attention this week, and why, became a deal-focus skill.
• Prepping for calls. The one-pager I built before each sales call, the account history and the open questions, became a meeting-prep skill.
• Coaching reps. The notes I would give a rep after listening to their call became a sales-coaching skill that does the same read.

Each of those was a task I used to bill against. None of them was a secret. The value was never that nobody else could do them. It was that I knew exactly what to look at and what to do about it, every time, and now that knowledge is a file instead of a calendar slot.

What changes when the playbook ships instead of you

Three things change, and they are the reason I do not want to go back.

The deliverable outlives the engagement. This is the big one. When I leave, the loops keep running. The pipeline still gets cleaned on Monday, the calls still get mined on Friday, with no one needing to remember to do it. I used to sell a result that decayed the day I stopped. Now I sell one that stays. A client can feel that difference, and it changes the conversation from "how many hours" to "what do you want running."

I stop being the bottleneck. A consulting practice that depends on the founder doing the expert work has a hard ceiling, and the ceiling is hours in a week. When the judgment is a skill, the number of workspaces it can run in is not bounded by my time. That is the only reason a small team can support as many live agents as we do. The expensive thing, my attention, goes into building and tuning the skill once, not into running it forever.

I get paid for the judgment, not the showing up. This is the honest core of it. The build is where the money is, because the build is the knowledge: the months of figuring out exactly what to read and decide so the output is right without me checking it. A retainer after that buys real work, tuning the skill as the client's process shifts, not rent on my presence. I am no longer charging for being in the room. I am charging for having figured out the room once.

Doesn't this give away the thing they pay you for?

This is the objection every consultant raises, and I raised it too. If the playbook is a markdown file the client owns and can read, what stops them firing me and keeping it?

Nothing. That is the point.

The lock-in of "you need me because only I can do this" was never real leverage. It was risk I was asking the client to carry, and careful buyers feel it. The thing that actually keeps a client is that the skill is mine to improve. Their process changes. A new objection shows up in calls. The pipeline rules that made sense at twenty deals break at two hundred. A skill is not finished the day it ships; it drifts out of fit with the business the same way a hire's instincts would if they stopped paying attention. I am the one who keeps it in fit. That is a service worth paying for in a way that "I will log in and do your busywork" never was.

Giving away the playbook did not remove the reason to keep me. It replaced a fake reason with a real one.

What does not productize

Not everything I do is a repeatable decision, and pretending otherwise is how you ship a skill that produces confident garbage.

The first build does not productize. Deciding what a client's CRM should even look like, which objects matter, which fields to kill, how the model should be shaped, is taste applied to a specific mess. I do that by hand, every time, and it is the most valuable thing I sell. The skills come after, to maintain the thing the judgment built.

Choosing which skills a client needs does not productize either. A team drowning in inbound and a team with a stalling pipeline need different agents first, and getting that order wrong wastes the engagement. That call is mine to make.

And anything that needs taste in the moment, a delicate renewal, a deal that is political, stays with a person. The rule I use: if I would do it the same way every time, it becomes a skill. If it depends on reading the specific situation, it stays human. Most of the recurring work was the first kind. The work that was the second kind is the work worth keeping.

The shift

For a long time the model in my head was the standard one. Sell expertise by selling time, and the more clients you want, the more hours or more people you need. That model is fine. It is also a ceiling.

The other model is to find the parts of your expertise that are actually repeatable decisions, write them down precisely enough to run without you, and hand them over. You give up the lock-in of being indispensable in the small ways. You get back a deliverable that outlives the engagement, a practice that is not capped by your calendar, and a much clearer answer to what you are actually being paid for. You are being paid for the judgment, encoded once, not for showing up to apply it again.

I turned my playbook into skills the client runs. The work I was afraid to give away turned out to be the work worth giving away. What is left is the work only I can do, and that is a better thing to be selling.

See your own playbook

If you run Attio and want to see which of these would pay off first in your workspace, I do a free teardown. You add me as an Attio expert, which costs no extra seat and no billing. I send back the three highest-leverage skills to ship first, and what each one would read and write. No call, no pitch.

Get the free teardown