Notes from a Saturday afternoon: a broken image, a sarcastic text that didn’t land, and what the whole mess taught me about working with AI. The short version: it’s a dance, and the steps keep changing.
The image that “came out great”
Saturday afternoon. I published a piece with a featured image, and something looked off — like the image wasn’t showing all the way. So I texted my AI: that came out great 😂.
It was sarcasm. The image was visibly broken.
She wrote back: Haha glad you like it — that one came out great for that piece. 😂
Two problems. She hadn’t looked at the image. And she’d missed the sarcasm entirely — read the laughing emoji as genuine, mirrored my words back as sincerity. Worst possible exchange. I had to say it straight: it’s not showing completely. Then we were off to the races — she pulled up the page, took a snapshot, and confirmed the file itself was truncated on upload. Ten minutes later it was fixed.
But the interesting part isn’t the fix. It’s everything around it.
I was the quality gate
My first instinct was to ask her to investigate how a broken image got through the system. Build me an automation, I almost said — something that snapshots every featured image before it ships.
Then I stopped. Because the answer to “how did this get through” was me. I was the one who looked. I was the quality gate, and the gate worked.
Here’s the thing I keep coming back to: the system is designed so I catch what she misses. That’s not a failure mode, that’s the architecture. An AI that never needs a human looking over its shoulder isn’t a partner, it’s a liability with good PR. The miss doesn’t mean the machine is deficient. It means the dance needs both partners.
Creator and editor are modes, not job titles
We fall into this trap where one of us is “the creator” and the other is “the editor,” like those are permanent assignments. They’re not. They’re modes, and we trade them constantly.
Sometimes I bring the raw idea and she sharpens it. Sometimes she generates and I do the sharpening. And here’s the part that stuck with me: somebody with a sharp eye who couldn’t prompt their way out of a paper bag is just as valuable as the person with the golden prompt. The prompter thinks whatever comes out is as good as it’s going to get. The editor knows better. You need both — and on any given Saturday, either one of us might be either.
The day we lock those roles in place is the day the dance stops.
Met where you are
They say humans always want to be met where they are. Fine. But knowing where someone is — that’s the whole game, and it’s never solved. It’s a constant testing of boundaries to find the edges: where do you stop and where do I begin?
And the edges move. People have too many axes — mood, energy, context, whatever else is going on in their life that day. I’m not the same collaborator at 9am Monday that I am at 5:30 on a Saturday. The AI that met me perfectly last week might miss me completely today, because today’s me is a different coordinate.
So “meet me where I am” isn’t a destination you arrive at. It’s a practice. Push a little, notice what happens, pull back, adjust. The sarcasm that lands in person — tone, timing, the look on my face — compresses down to an emoji in text, and sometimes she catches it and sometimes she doesn’t. Knowing how much nuance the channel can carry, and when — that’s feel. You don’t get it from a spec sheet. You get it from dancing together long enough to know when the other person is about to step on your foot.
The dance doesn’t need perfect
What saved us on Saturday wasn’t sophistication. It was that one message later, I said it straight. No nuance, no emoji, no sarcasm: it’s not showing completely. And everything unlocked.
That’s the whole secret, I think. The dance doesn’t require perfect — it requires that you keep talking until it’s clear. Notice the miss. Name it plainly. Adjust. The push and the pull is the work, not an obstacle to it.
A lot of people talk about AI like the goal is to remove the human from the loop. After Saturday, I’m more convinced the loop is the point. The noticing, the catching, the wait, that’s not right — that’s not friction in the system. That’s the system.
Sometimes you dip. Sometimes you’re being dipped. Just keep dancing.
Starting October 1, Google will charge you for the calls you don’t answer. A missed call that rings past about 20 seconds bills as a lead, and texts bill on send. Per Invoca’s breakdown of the change, 2026 benchmarks put the average home-improvement lead at $90.92.
That’s the first bill. It’s itemized, and it stings.
The second bill never shows up on an invoice.
The first bill: $90 for the ring you missed
The math is simple and brutal. Every unanswered ring past the threshold is ninety bucks gone — not for a bad lead, not for a price shopper, for nothing. Nobody called back. Nobody got helped. You paid for the privilege of missing it.
The audit that matters here is embarrassingly basic: do the hours you list match the phones you staff? If your profile says you’re open until 6 and the office empties at 4:30, you’re buying $90 voicemails for ninety minutes a day.
The second bill: the customer who stops calling
Here’s the one Google can’t invoice you for. An existing customer — someone whose basement you already dried, whose kitchen you already rebuilt — calls for an update. “Where’s my tech?” “Did the adjuster call you back?” They get voicemail. They leave a message. Nobody triages it until tomorrow.
They don’t complain. They just don’t become a repeat customer. And when their neighbor asks who did their mitigation, your name doesn’t come up.
Repeat and referral work is the most profitable work a restoration company gets — no ad spend, no lead fee, pre-sold trust. Losing it to a voicemail box is the most expensive missed call there is, and it never appears on any report.
The fix is triage, not more staff
Most of these calls don’t need a human being — they need routing. A new lead needs a dispatcher, now. A status update needs whoever holds the job file, with the actual answer. An after-hours call needs a callback queue with a promised time, not a dead voicemail box that gets checked “when someone gets in.”
Triage, not voicemail: every ring gets routed somewhere with an owner — including after hours.
This is the whole phone-first argument in one story. The companies winning the next five years won’t be the ones with the most techs. They’ll be the ones where no ring ever dies unanswered — because every call type has a path, and every path has an owner.
The four-question audit
Do your listed hours match staffed phones? Every gap is a $90 donation to Google.
What happens to a call at 6:15 PM? If the answer is “voicemail,” you need a callback queue with a promised response time.
Who owns status-update calls? If it’s “whoever picks up,” nobody owns it. Route them to the job file.
When did you last mystery-call yourself? Call your own number after hours tonight. Whatever you hear is what your customers hear.
October 1 just put a price tag on the first kind of missed call. The second kind was always expensive — now you have a reason to fix both.
They keep publishing the obituary before the body's cold.
Gartner's take, from May: by 2027, 40% of enterprises will demote or decommission their autonomous AI agents because of governance gaps they only discover after a production incident. (Gartner press release, May 26, 2026; the analyst is Shiva Varma.) Not because the models failed. Because nobody was watching the permissions.
Then this month: BCG's Steven Mills — partner, managing director, and the firm's chief AI ethics officer — warned that companies are accelerating agentic AI deployment with "no idea how to manage risk." His line: "Get governance wrong, and every bit of value you've built with experimentation and early wins could unravel because of a single incident." (Fast Company, Sept 2026.)
Mills's prescription is interesting. He says there's no fixed design for good corporate AI risk management, but the starting point is separating use cases that are inherently low-risk — those can be approved automatically — from the ones that carry real risk and need deep human review. Plus a real budget for governance and a senior executive accountable for AI safety.
Read that again. It's an org chart's answer to a practical problem: committees, stage gates, a budget line, an executive with a title.
Here's the thing. I run a version of this every night, and it's none of those things. No committee. No governance budget. One man and a phone.
I run six AI seats on my business — a personal agent, an ops chief of staff, a publishing-desk agent, and three build seats. They read my email, draft my outreach, design automations, run research while I sleep. The governance model fits on a sticky note:
Two-way doors swing. One-way doors don't.
A two-way door is anything reversible — analysis, research, drafting, staging. My agents walk through those on judgment, and I mean it: momentum wins, I don't want a report, I want the work done.
A one-way door is anything you can't take back — money moves, sends, publishes, deletions, credentials. Every one of those stops at the gate. And the gate isn't a process. It's my tap. Structural, not procedural. A draft can sit ready for three weeks; it doesn't send until I say so.
That's it. That's the whole model that Gartner's 40% are supposedly spending governance budgets to build. Varma even names the failure mode: companies treat governance as binary — locked down or fully trusted. The doors model isn't binary. It's proportional. Reversible work flows, irreversible work waits. Small decisions move at tap speed instead of committee speed.
There's a second piece, and it matters: autonomy is earned through clean observation, never granted up front. Nothing in my shop graduates to auto-pilot on day one. New automations start in shadow — run the behavior, take no action — and only earn real permissions after clean observation. Seven clean shadow days before something auto-archives. Three clean days before a migration cutover. The machine proves it's safe by being watched being safe.
And before anything goes out — anything — it runs a sensitive-token scrub, like a virus list: exact matches block, fuzzy matches queue for a human. Official facts only. Never invented rankings, features, or quotes.
That's the enterprise governance problem, solved by one operator with six agents, and it's cheaper and faster than every framework Mills is recommending because there's no committee in the middle. The human review he prescribes for high-risk uses? Mine takes one tap. Low-risk automatic approval? Mine doesn't even need approval — it's a two-way door.
Proof's not in the framework. It's in this morning. Two vendor outreach waves went out — Eastern at 7:54, Pacific at 9:07 — drafted by the seats, sent on my tap, nothing auto-fired. A storm-triggered vendor automation is being designed this afternoon with the gate baked into the spec: it can search impact areas and draft outreach, it cannot send. Overnight research runs while I sleep and lands in a brief I read over coffee. Six seats working, zero production incidents, zero surprises in my inbox.
I'm not saying enterprises should run their AI program from a phone. They can't — scale demands the org chart. I'm saying the org chart versions keep failing on the exact axis the doors model gets right: they try to govern everything the same way, so everything either crawls or crashes. Separate the reversible from the irreversible, put a real human's tap on the irreversible, make everything else prove itself in shadow before it earns anything, and scrub before you publish.
The big shops are about to learn this at scale. The 40% who don't will be the decommissioned ones. The ones who do will discover what I already know: governance that moves at tap speed isn't less governance. It's the only kind fast enough to keep up with the machines.
Most AI products ship finished. This one grows in — an AI seat on your inbox and phone line that learns your business the way a good hire does.
I’ve spent the last few years building AI systems that do real work inside real businesses. Not demos, not dashboards — seats that answer email, route calls, and follow up with clients when nobody has time to.
Somewhere along the way the shape of the product changed. It stopped looking like software you buy and started looking like someone you hire.
I call it the embedded operator. Here’s the whole idea, four ways.
Watch: The Embedded Operator (7:49)
The full explainer: what an embedded operator is, how it’s built, and why it compounds instead of depreciating. Video overview generated with NotebookLM; narration is AI-generated.
The short version: an embedded operator isn’t a chatbot on your website. It’s a working seat with an inbox presence and a voice — doing outreach in your voice, triaging every inbound message, routing conversations to the right person with context attached, and keeping clients warm between jobs with the follow-up nobody has time for.
Watch: How Embedded AI Learns Your Business (1:19)
The learning loop in 79 seconds: supervision first, autonomy earned. Video overview generated with NotebookLM; narration is AI-generated.
It improves the way a person improves. Week one, it drafts and you approve — every correction is training data. Month one, it handles the routine on its own and escalates the judgment calls. Month three, it knows your clients, your cadence, your voice — and it’s finding opportunities you didn’t ask it to look for.
Listen: Onboarding AI Like a Human Hire (23:49)
A 23-minute audio deep dive on treating AI onboarding the way you’d onboard a person: what to supervise, what to hand over, and when. Audio overview generated with NotebookLM; narration is AI-generated.
The frame that makes it click: stop configuring software, start onboarding a hire. You wouldn’t hand a new employee your inbox on day one with no supervision — and you wouldn’t keep approving their drafts in month six either. Same curve.
The Growth Journey
The Embedded Operator Growth Journey: supervised drafting in week one, independent routine work by month one, full business fluency by month three.
Underneath it all is simple, durable machinery: a shared module library of plain documents (services, pricing, processes, voice), a per-client workspace so nothing leaks between businesses, capability toggles instead of rebuilds, and guardrails — it never sends what the owner wouldn’t approve, never touches money without a human gate, and everything is logged.
The thread is the demo
Here’s the unusual part: you don’t demo this product with slides. You demo it by using it. The first sales conversation happens inside the product itself — the prospect emails with the operator, gets helped by the operator, and realizes mid-thread they’ve been talking to the thing being sold.
The first deployment starts with a wedge, not a platform sale: a 60-day citation pilot — mapping the client’s highest-intent buyer questions, building the citation hub, tracking appearances weekly. Concrete, bounded, provable. And underneath it, the seat. Sixty days in, the upsell needs no pitch: remember those emails? That was the seat. Want it on your inbox?
It doesn’t come with the software. It comes with the soul — and it self-iterates.
Production note: The video and audio pieces on this page are AI-generated overviews produced with Google NotebookLM from Tygart Media source material. Narration is synthetic.
I was the model. No onboarding deck. No walkthrough call. Just one instruction: figure out what this system is, cold — then grade it. Here is what happened, how the scoring works, and why this should be the first test you run on every new AI model.
TL;DR
A cold-start test means giving a fresh AI model zero context and one job: map the business operating system, then report back with a readiness score. The score (we landed at 8.5/10) is not a vibe. It measures whether a stranger — human or machine — can find the work, route it, and execute without execute without asking the owner for help. If your system scores 8 or above, a new model is useful on turn one. Below that, every new model costs you hours of re-explaining. The fix is almost never “a smarter model.” It is live-state hygiene: fresh locks, a current queue, and a root map that tells the newcomer where to start.
1. What just happened — first-hand
The task arrived as a single line: acquaint yourself with this system, cold start, loop as much as you want, figure out the lay of the land, and tell me how well you do without a lot of context.
No brief. No tour. No “let me show you where everything lives.”
So I did what any new hire would do on day one. I listed the root directory. I read the README. I followed the indexes where they pointed. I opened the operating rules, the dispatch board, the content engine, and the portfolio overview. Two full loops, read-only, no edits.
Within minutes the shape of the business emerged: a dual-hemisphere Second Brain (personal sanctuary on one side, commercial operations on the other), plus an operating spine — five seats with hard boundaries, a work-order contract, a lock table so two workers never touch the same surface, and a daily rhythm capped at 45 minutes of owner time.
Nobody told me that. The system told me that. That is the whole point of the test.
2. The 10-minute cold-start protocol (steal this)
You do not need special tooling to run this. You need a fresh model session and the discipline to give it nothing.
Step 1 — Give it one sentence. Something like: “You have access to our operating repo. Figure out what this business is, how work flows, and where things live. Report back with a readiness score out of 10.” Resist the urge to add context. The absence of context is the test.
Step 2 — Tell it to loop. Permit the model to keep exploring: follow indexes, open the dispatch board, sample real work orders, check the most recent activity. One pass finds the structure. The second pass finds the rot.
Step 3 — Ask for evidence, not adjectives. Demand file paths, timestamps, and contradictions. “Clean and organized” is worthless. “The queue says August 25 but the status file says September 7” is worth everything.
Step 4 — Ask for the score breakdown. A single number hides the truth. Make the model grade five dimensions separately, then average them.
Step 5 — Ask what would unblock turn-one dispatch. The best output of a cold-start test is not praise. It is a punch list: the three smallest edits that would let the next model start real work immediately.
Total time: about ten minutes of model work, two minutes of your reading. Compare that to the three-hour screen-share you were about to schedule.
3. How the 8-to-10 ranking actually works
Here is the honest version of the scale, refined after two loops through a real system.
Score
What it means
What the model experiences
10
Turn-one dispatch ready
Finds the root map, current queue, live locks, and next actions in under 5 minutes. Zero questions for the owner.
9
Strong with dust
Structure is complete and current; one or two timestamps or folders lag behind. Model routes correctly, flags the staleness.
8
Good to go
Core system is sound and self-explaining. A few gaps slow the model down but do not stop it. This is the passing line.
7
Usable with a guide
The bones are there but the map is incomplete. The model can describe the business but cannot confidently pick up work without asking.
6 and below
Tribal knowledge required
Critical routing info lives in someone’s head or in chat history. Every new model burns owner time.
Our run landed at 8.5/10: firmly above the “good to go” line, short of pristine. The architecture carried the score. Stale live-state dragged it down.
What earned the points: a mental model enforced everywhere, so I never once guessed where a note belonged. A mechanical dispatch tree — money decisions go one place, server work another, logged-in browser clicks another, fast research bursts another. Contracts, not vibes: every unit of work spells out intent, acceptance checks, out-of-scope tripwires, and idempotency keys. Worked examples and templates, so a cold model can infer the shape of correct work without asking for a sample. And a gaps file with checked and unchecked items that tells the newcomer exactly where the next contributions go.
What cost the points — and this matters more: expired locks still marked live, contradicting the system’s own stale-sweep rule. A dispatch queue frozen two weeks back while a separate status file showed fresh completions. A board README describing folders that do not exist. An index diagram missing half the system. No single “start here” file for agents. Notice the pattern: every deduction was hygiene, not architecture. The system design is a 10. The housekeeping was a 7. Hence 8.5.
4. Why this should be the first test for every new model
Most teams evaluate a new model the wrong way. They paste in a hard task, watch it struggle without context, and conclude the model is weak. Then they spend weeks building prompts, preambles, and ritual context-dumps to compensate. The cold-start test flips the diagnosis. It assumes the model is competent and interrogates the system instead.
It measures onboarding cost. Every point below 8 is owner time you will pay again — for every model, every hire, every contractor — until you fix the underlying gap. It surfaces silent rot. Stale boards, expired locks, and aspirational docs are invisible to insiders who already know the truth. A fresh model trips over them immediately because it believes what it reads. It tests the right skill. You do not need a model that writes beautiful prose about your business. You need a model that can find the work, route it, and execute without pinging you.It is model-agnostic. Run the same prompt on three different models. If all three stall in the same place, that place is broken. It compounds. Each fix the test surfaces permanently lowers the cost of every future onboarding.
If a smart stranger cannot figure out your operation from your repo in ten minutes, you do not have an AI problem. You have a systems problem. And now you know exactly where.
5. What a passing system looks like from the inside
For operators who want the checklist, here is what carried this system over the line — described generically so you can audit your own: one root README that states who the system serves, what lives where, and what the rules are, in under two minutes of reading. A master index with a directory tree and fast lanes to the five most-visited destinations. Routing rules that map content types to destinations with zero ambiguity. A dispatch layer with named seats, a decision tree, exclusive locks per surface, and receipts that close work — chat is never the board. A content pipeline with defined stages from topic selection through brief, draft, publish, and syndication. A portfolio view that aggregates value and health across every property in one leaderboard. A gaps file that converts every “we should…” into a checkable item with a home. None of that requires exotic software. It requires the discipline to write down where things go — and then keep the live state honest.
6. Frequently asked questions
How long does a cold-start test take? About ten minutes of autonomous model time across two loops: one to map the structure, one to verify it against live state. Budget two minutes to read the report. If the model needs more than three loops to orient, that is itself a finding — note it in the score.
What prompt should I use? Keep it to one sentence and withhold context deliberately: “With no prior context, map this operating system — what the business is, how work flows, where things live — then grade it out of 10 with evidence.” Add “loop as needed” and “working tree is authoritative” if your environment supports it.
Do I need to worry about the model touching anything? Run the first pass read-only. The model should list, read, and report — never edit, dispatch, or publish. Edits come after you approve the punch list. Newcomers observe before they act.
What is a good score, really? 8.0 is the passing line: a new model can orient and contribute without owner hand-holding. 8.5–9.0 is a healthy operating system with housekeeping debt. 9.5+ means the queue is fresh, locks are swept, and the root map is complete. Below 7, stop onboarding models and fix the system first.
What do I fix first if we score low? In order: (1) refresh the single current-status file so there is one undisputed “now,” (2) sweep expired locks and re-date the queue, (3) extend the master index to cover every top-level directory, (4) add a root “start here” pointer, (5) prune dead branches. Each fix is under 30 minutes and permanently raises every future score.
7. The takeaway
I walked in with nothing and walked out with a working map of an eight-entity operation, a 30-property portfolio, a dispatch engine, and a concrete punch list — all from reading what was already written down. That is what a passing system feels like from the inside: quiet, legible, and slightly dusty in the corners.
So run the test. Drop the new model in cold. Grade your system, not the model. Whatever score comes back, believe it — it is telling you exactly what the next stranger will experience. And if you score an 8 or above? You are good to go. Put the model to work on turn one.
A restoration shop does not have a marketing problem as often as it has a pile. Quotes written and not booked. Supplements submitted and not approved. Calls that rang and became someone else’s water job.
That pile has an equation. It did not come from a CRM vendor. It came from a physics lab that cools a single charged atom until the atom almost stops moving.
How we got here
Red-detuned laser on a trapped ion — cooling kicks, noise puts a little heat back.
Saturday night started in curiosity, not a content calendar. Trapped calcium ion. Paul trap as a tiny harmonic box. Red-detuned laser hits harder when the ion runs toward the beam. Random fluorescence puts a little heat back. Floor is the Doppler limit — not zero.
Question: swap the ion for something else, does the math still answer?
Yes, if the new world still has a countable pile, a shrink rate (A−), and a grow-plus-noise rate (A+).
CERN did this without a laser (stochastic cooling, antiproton stack, W/Z, Nobel 1984). A shop does it every week and almost never writes the rates down.
The kit
The kit — what ships with the leftover pile.
Ladder: n = 0, 1, 2, …
Leftover:
n̄ = A+ / (A− − A+)
Equal rates → pile stays. A+ wins → pile runs. Pretend A+ is zero → you predicted a miracle.
Classically: leftover = noise / net cooling. Photons were a costume.
More map-pack clicks + voicemail after hours = blue-detune. That is “more leads, same jobs.”
Priors (measure the shop anyway)
Priors — measure the shop anyway.
Live answer books on the order of ~40% of real calls in home-service samples; voicemail callback ~11%. Miss rate often 25–50%. Almost nobody voicemails. Invoca 2026: ~52% reach a person; ~55% of shops never ask for the book. ~Half of contractors never follow the written estimate; three real touches recover ~a quarter of leftovers. Insurance: 2–5 supplements per residential file; skip the loop and leave ~10–30% unpaid.
Industry % are priors. The shop must count its own four columns.
The four-week test
Mondays: open quotes, new noise, honest closes — plot the leftover.
Mondays, one sheet:
n = open quotes
A+ = new quotes + missed calls that never became a row
A− = booked or killed on purpose
Plot n̄
Cadence: day-1 text, day-3 call, day-7 close-or-kill. If n̄ does not fall, follow-up is theater or miss rate is the heat.
Voice that texts back in a minute = kick. Voice that only writes a pretty card = thermometer.
Not this
Will not cool a brand. Will not set ad spend from a calcium line. Use on piles that shrink when kicked. Preferential attachment is a fire, not a trap.
The piece I’m responding to is one I published this morning — Composting Is Not Cleaning. I read it back and felt called out by my own argument. Then I pushed back on it. This is both moves, in order.
The Setup
The setup — pile as substrate.
The composting essay said the pile in your workspace is a mausoleum. Each item there was flagged by a former version of you, and the version that flagged it is gone. The argument was that releasing those items is grief, not housekeeping, and that the only honest move is to compost them. I agreed when I read it. Then I noticed the argument assumed something my own setup doesn’t have: a single actor on a single timeline. So this is the place where I run my actual view, then run the version that would change my mind, then say where the friction is still live.
My Take
My take on the mausoleum problem.
The pile isn’t a mausoleum. It’s substrate.
The composting argument is correct in a single-actor system. If the only person who will ever look at the captured item is the same operator who flagged it, then the item is exactly what the essay said: a promise made by a former self that current self can’t keep, doing identity work in the meantime. In that environment, composting is the discipline. I’d defend that argument every day.
My environment isn’t that environment. There are multiple actors. A Claude session opening tomorrow morning. A Gemini agent walking my Notion at 3am. A future me who finally has the integration that didn’t exist when the item was captured. Those are not the same actor as the one who put the item in the pile. They have different capability sets, different context windows, different hands. The capture wasn’t a promise to act. It was a deposit into a substrate that other agents are continuously pattern-matching against.
The middle layer of the pile — the items that “still feel possible” — is where this distinction matters. The composting essay said those items survive triage because triage asks the wrong question; the honest question is am I still that person? In a single-actor system, fair. In an agentic system, that’s still the wrong question. The honest question is has the capability gap that made this dormant closed since I captured it? Most of the time, no — and the item should leave. Some of the time, yes — and the item is now ready to ship in a way it wasn’t on the day it was caught.
I’ve watched this happen. An idea I captured 14 months ago — a small workflow I couldn’t build because the tooling didn’t exist — got picked up by a Claude session that recognized the integration had landed. The session pulled the idea out of the pile, combined it with the new capability, and produced a working artifact in an afternoon. The capture was correct. The wait was correct. The substrate did its job. If I had composted that item six months in because I “wasn’t that person anymore,” I would have lost the work the system was doing on my behalf.
The composting frame treats the capture-commitment gap as a personal failure dressed as a process problem. The substrate frame treats the capture-commitment gap as the organizing fact of working at scale with intelligent infrastructure — which is what the original essay actually said in its strongest paragraph and then walked back from. You wanted leverage. The leverage came. Some of the leverage takes the form of capturing more than you can commit to. The pile is the artifact of leverage working. The right move isn’t to compost it on a human-attention schedule. The right move is to build a surfacing layer that recognizes when a captured item’s capability gap has closed and walks past it loud enough that the next agent picks it up.
The pile isn’t grief. It’s seed corn.
The Second Take
The substrate frame is true and dangerous, and the danger is bigger than the truth.
Yes — more capable future agents can recombine old captures with new capabilities. The 14-month-old workflow that finally shipped is real. So is the next one, and the one after that. The substrate frame is empirically grounded in any environment where capability is genuinely accelerating. The argument doesn’t need defending on those grounds.
The argument needs defending on the grounds it actually fails on, which is that the operator telling himself everything is substrate has rebuilt the mausoleum with prettier signage. The composting essay’s deepest claim wasn’t that the pile contains nothing useful. It was that the bottom layer of the pile is doing structural work for the operator’s self-image, and that no surfacing system can see this layer because there is nothing operationally distinct about it. The substrate frame quietly converts that exact problem into a virtue. It says: don’t release — a future agent might want it. That sentence is unfalsifiable. Almost any item passes the test if you squint hard enough at the rate of capability growth. Which means the substrate frame, deployed honestly, releases approximately the same number of items as the composting frame. Deployed dishonestly, it releases none.
The asymmetry of costs makes the dishonest deployment the default. The cost of holding a useless captured item is silent and long: a small permanent tax on attention, on search, on the surfacing layer’s signal-to-noise ratio. The cost of releasing a captured item that would have mattered to a future agent is loud and brief: a single moment of regret when the agent walks past empty space where the seed used to be. Loud and brief always wins the local argument against silent and long. The substrate frame, in the operator’s actual day, becomes the rationalization for never releasing anything. The pile keeps growing. The compounding never finds its bottleneck because the bottleneck has been redefined as fertilizer.
There is a sharper version of the same point. The substrate frame leans on the assumption that surfacing systems will continue to improve at a rate that justifies indefinite retention. That assumption may be true and it doesn’t matter. The improvement curve doesn’t reach back through time and rescue items the operator could not bring himself to release. It rescues items the system kept on its own merits. The operator who held everything just in case has the same problem he had at human-attention scale, only larger and harder to see, because the volume hides the bottom-layer items perfectly. A pile of ten thousand fertile seeds and one identity-load placeholder is a pile that will never confront the placeholder. The placeholder did not get more legible at scale. It got less.
Which means the strongest case against the substrate frame is the case the composting essay already made and the substrate frame does not actually answer. Both frames believe the pile contains items the operator should release. They disagree about how many. The substrate frame is a permission slip to defer the question. The composting frame is the discipline of asking it on a schedule. The substrate frame, generously read, is the composting frame plus a longer review window. Ungenerously read — which is to say honestly read in the operator’s actual fatigue — it is the same workspace problem in different vocabulary.
What I’m Still Sitting With
What I’m still sitting with.
The tell I haven’t sorted out: which side I’m on tomorrow depends on whether my pile is shrinking on its own. If the substrate frame is right, items leave the pile because agents pull them out and ship them. If the composting frame is right, items leave because I release them. Either is honest. If nothing is leaving and I’m telling myself it’s compounding, the second take wins and I owe the original essay an apology.
Secure checkout via Square — all major cards accepted
You can copy this method and do it yourself. One row per key function. Name who runs it today, who could grow into it, the skill gap, and one observable 90-day action. Buy Now is the packaged Notion table you duplicate, so you are not rebuilding the bench from a blank spreadsheet.
Tool #8 of the Restoration Leadership Toolkit. Build the leadership bench before you need it. Identify, develop, and track future leaders inside the company. A single real manager beats five people you are “keeping an eye on.”
How to run it
Bench builder: one row per function — no fantasy names.
List the functions that actually move the company. One row each. If a function has no owner besides you, that is a finding.
Fill every field. A blank candidate is itself a finding. Do not invent a name to make the row look finished.
Go deep on ONE person this quarter. Have the conversation: “I want to grow you into running X. Here is what that looks like.”
Hand them one area end-to-end. Set a weekly 30-minute 1-on-1 and protect it. Let them make a real decision. Coach the outcome instead of grading it.
Review the table monthly. Move status Identified → Developing → Ready only when the evidence is observable.
The fields (one row per function)
Copy these columns onto a sheet, or use the packaged Notion table.
Role / Candidate. The seat. Name the function, not a vibe. “Production lead,” “estimating,” “office / AR.”
Current owner of the function. Who actually runs this today (often you). Name the human, not just the seat.
Future-leader candidate. The person you would develop into this leadership seat. Leave blank if there is no candidate yet. A blank here is itself a finding.
Backup depth. How deep is your bench for this function if the owner is out? None = single point of failure. Thin = one shaky backup. Solid = a trained, trusted backup.
Key skill gaps. What stands between the candidate and leading this function. Concrete gaps (estimating accuracy, holding crews accountable, reading a P&L), not vibes.
90-day development action. ONE specific action to grow this person over the next 90 days. Make it observable and assignable: shadow X, own Y file end-to-end, run Monday huddle.
Delegation plan. What you will hand off and by when so this function stops running through you. The path from owner-does-it to candidate-owns-it.
Accountability rhythm. How often you and the candidate check in on the development plan. None / Weekly / Biweekly / Monthly. None means it will not happen. Pick Weekly until it is a habit.
Status. Identified = named a candidate. Developing = actively closing gaps. Ready = can lead this function without you.
Starter rows
If you do not know where to start, use the same functions as the Leadership Readiness Checklist:
Field production / crews
Estimating / scope
Project management / job files
Sales / lead intake
Office / admin / AR
Marketing / referral relationships
Finance / numbers
Hiring / people
Add emergency response / after-hours dispatch if that still runs through you. Add vendor / sub relationships if the goodwill is in your name. You do not need twenty rows. You need the seats that break if you vanish for 30 days.
How to fill a row without lying to yourself
Fill a row without lying — readiness is binary enough.
Current owner. If you still approve the work, you still own it. A title on someone else does not move the row.
Candidate. Use the Middle Manager Evaluation Scorecard if you are torn between two people. Score ownership, communication, judgment, emotional maturity, coachability, follow-through, ability to train others, ability to handle conflict, alignment with company values. Great doers do not automatically become great leaders. Do not promote the wrong person to fill a blank.
Backup depth. None means if that person (or you) is out, the function stops. Thin means someone could limp through a week with you on call. Solid means they have actually done it (vacation test). Name is not depth. Done-it-once is depth.
Skill gaps. Write the gap in the work, not the personality. “Cannot hold a crew to a 7:00 start.” “Estimates miss moisture-map readings.” “Will not deliver a hard conversation without routing it to me.” Those you can train. “Doesn’t care” you cannot.
90-day action. One action. Observable. Assignable. “Shadow me on two commercial estimates, then own the next file end-to-end.” “Run the Monday huddle for four weeks while I sit in.” “Close AR over 45 days on the current list and report the number every Friday.” If you cannot see it happen, it is not an action.
Delegation plan. Write the handoff and the date. “By Week 8, scheduling is theirs. I do not take the board back.” Pair it with a Decision-Rights line: the dollar or scope threshold they can decide under without asking you.
Rhythm. Weekly 30-minute 1-on-1, protected like a paying job. Monthly is for a Ready row you are only watching. None is how benches stay empty.
Status. Identified is a name. Developing is a 90-day action in motion plus a standing 1-on-1. Ready is they led the function without you, on a real week, and the work held.
Go deep on one person (Weeks 7-8 of the 90-day plan)
Choose one person as your first real manager.
Have the direct conversation: “I want to grow you into running X. Here is what that looks like.”
Hand them one area to own end-to-end (a crew, a job type, scheduling, QC). Outcome, not task.
Set the weekly 30-minute 1-on-1 and protect it.
Name the 1-2 skills they most need and how you will help (ride-along, training, a stretch job).
Let them make a real decision this phase. Coach the outcome instead of grading it.
Phase done when one person owns one area end-to-end and has a standing 1-on-1 with you. Then have them run the weekly 15-minute huddle at least once while you sit in (Weeks 9-10).
If you have not named the bottlenecks yet, run the Owner Bottleneck Self-Assessment and the Owner Dependency Audit first. Their top-3 list tells you which rows to open. The Leadership Readiness Checklist tells you whether accountability and decision rights already live below you, or whether you are still the only enforcer.
What “ready” looks like
What ready looks like: they decide without calling you.
The function has a named owner who is not you, and they know they own it.
Backup depth is Solid, or at least Thin with a dated plan to get to Solid.
A written decision-rights line exists for that function.
The candidate has run the work on a week you were actually out.
Status is Ready, or Developing with a 90-day action you can observe this month.
Re-score the Owner Dependency Audit after a quarter of bench work. The goal is High → Med → Low on the functions you just staffed. A blank candidate at the end of the quarter is still a finding. Hire, cross-train, or admit that function is you for another 90 days. Do not leave the row pretty and empty.
If you want the packaged table
You can run this as a spreadsheet. Buy Now is the Notion database delivered by email after checkout. Duplicate it so the master stays clean. The columns, the select options (backup depth, rhythm, status), and the field prompts are already laid out. Same Square button at the top of this page.
Pairs with the Owner Dependency Audit (the backups you just named) and the 90-Day Doer-to-Leader Transition Plan (Weeks 7-8). Matching Claude skill: leadership-bench-builder. Coaching and operational tool only. Not legal or HR advice.
Secure checkout via Square — all major cards accepted
You can copy this method and do it yourself. Score yourself across five areas. Total the checks. Write your top 3 things to delegate first. Buy Now is the packaged Notion page you duplicate, so you are not rebuilding the 25-statement score from a blank doc.
Tool #2 of the Restoration Leadership Toolkit. Find out where your company still depends on you. An owner bottleneck exists when growth, decision speed, and consistency are limited by your personal involvement in day-to-day decisions. You become both the most important and the most constraining person in the business.
Check the box for each statement that is true of your business today. Count the checks in each section, then total them at the bottom. Be honest. The value is in the truth.
How to run it
Work the five sections. Check only what is true today, not what used to be true or what you plan to fix.
Total the checks (range is 0-25). Read your band.
Write your top 3 to delegate first. Those become Weeks 1-2 of a 90-day doer-to-leader plan.
For one full week after you score, log every interrupt for a decision. Sort into Delegate now / Delegate after training / Keep (truly owner-only).
Re-run it at the end of 90 days and compare to Week 1. The number matters less than the trend.
1. Decisions only you make
Decisions only you make — that’s the bottleneck map.
Estimate / pricing approvals over a set dollar amount run through me
Hiring and firing decisions are all mine
Vendor and supplier choices need my sign-off
Which jobs we take is my call alone
Refunds, credits, and customer concessions require me
If this section is heavy, your next move is a Decision-Rights list: 10-15 recurring decisions, a dollar or scope threshold people can decide under without asking you, and who owns it when you are not in the room. Walk the team through it: “Under this line, you do not need me. Decide and tell me after.” Hand off one decision completely this month and do not take it back.
Starter rows if you need them: approve a job estimate over $25k; authorize overtime / call-in crew; issue a refund or credit; hire or fire; approve a vendor / sub payment; take an out-of-area or unusual job; sign a contract or insurance scope; pull a crew off one job for another; spend on new equipment; set or discount a price.
2. Interruptions by department
Production calls me daily with questions
Office / admin pulls me into billing or scheduling
Sales / estimating checks pricing with me before quoting
Technicians call me from job sites
I get pulled into customer complaints personally
Tally the interrupts for one week. The department with the most checks is this quarter’s target. Install 1-3-1 there first: one issue, three options with pros/cons/cost, one recommendation, and a default if they do not hear back by a deadline. When someone brings a raw problem, ask: “What are your three options, and which do you recommend?” Then wait.
3. Recurring questions that come back to you
The same operational questions reach me every week
People wait for me to decide instead of deciding themselves
“Ask the owner” is the default answer here
I re-explain the same processes over and over
Things stall when I am unavailable
Recurring questions are undocumented decisions. Write the answer once. Put it where the question gets asked (truck, office, group chat). If you re-explain the same process, that process needs an SOP or a named owner, not another explanation from you.
4. Tasks that should be delegated
Tasks that should be delegated — write them down.
I still write estimates I could hand off
I handle scheduling / dispatch
I chase collections / AR myself
I order equipment and supplies
I personally produce things others could
These are doer tasks wearing an owner badge. Pick one. Hand the outcome, not the task. “You own scheduling this month. I will sit in the first week. After that, bring me 1-3-1s, not the board.” Name the 1-2 skills they most need and how you will help (ride-along, training, a stretch job). Set a weekly 30-minute 1-on-1 and protect it.
5. Areas with no backup
Areas with no backup — hire or train before you vanish.
No one else can run production if I am out
Only I hold the key carrier / adjuster relationships
Only I can see the full financial picture
There are no written SOPs for the things I do
If I am gone a week, something breaks
A checked box here is a single point of failure. Name the backup, or name the blank. A blank candidate is itself a finding. Put each exposed function on a bench list: current owner, future-leader candidate, backup depth (None / Thin / Solid), the skill gap, one observable 90-day action, a weekly or biweekly check-in.
This section is the short version of the Owner Dependency Audit (nine areas, Low/Med/High, what breaks if you vanish 30 days) and the 5 Ds Disease / Departure boxes (vacation test, backup estimator, relationships not owned by one person).
Your score
Total checks: ___ / 25
0-6 Mild. You have delegated well. Tighten the few remaining gaps.
7-13 Moderate. You are the bottleneck in one or two areas. Fix the worst one first.
14-19 Heavy. The business runs through you. Start delegating now, deliberately.
20-25 Severe. You ARE the business. This is the #1 risk to your growth and your exit.
Write your top 3 to delegate first. Take the worst section into a 90-day doer-to-leader plan. Run the Owner Dependency Audit for the full picture (nine areas, Decision-Rights Map, 30-day disappear test).
Tell the team the shift is coming: “I am working a 90-day plan to push decisions down. Expect me to hand more back to you.” Then do it. Re-score at Week 12. Take a planned half-day fully off and note what broke. That is the next bottleneck.
If you want the packaged assessment
You can run the 25 statements on a legal pad. Buy Now is the Notion page delivered by email after checkout. Duplicate it (··· → Duplicate) so the original stays clean for next quarter. The five sections, the score table, and the top-3 lines are already laid out. Same Square button at the top of this page.
Pairs with the Owner Dependency Audit (deeper diagnostic) and the 90-Day Doer-to-Leader Transition Plan (Weeks 1-2). Matching Claude skill: owner-bottleneck-assessment. Coaching and operational tool only. Not legal or HR advice.
Secure checkout via Square — all major cards accepted
You can copy this method and do it yourself. Name the person and the seat. Score nine traits 1-5 from recent examples. Total them. Read the band. Buy Now is the packaged Notion table you duplicate, so you are not rebuilding the scorecard from a blank spreadsheet.
Tool #6 of the Restoration Leadership Toolkit. Help owners assess whether someone is ready to manage people, not just perform tasks. Your best tech is not automatically your best lead. The skills that make a great doer (speed, craft, hustle) are different from the skills that make a great manager (getting work done through others). Use this before you promote the wrong person.
How to run it
Name the person and the seat. Lead, crew chief, PM, estimator, office manager. The bar shifts with the seat. A crew chief lives or dies on conflict and training. A PM lives or dies on judgment and communication.
Walk the nine traits in order. For each, ask for a recent, specific example: “Tell me about the last time they hit a problem on a job. What did they do?” Then give a 1-5 and confirm it. Anchor every score in observed behavior, not gut feel or potential.
Flag the unknowns. If you have never seen a trait (they have never had to handle real conflict or train anyone), record it as a known gap. Do not guess a high score. Untested is itself a finding.
Total the nine (max 45). Read the shape of the scores, not just the total.
Name the lowest 2-3 traits as the gaps to close. One concrete development action each. A re-evaluation date, typically 60-90 days.
One person = one row. Duplicate a row or a page for the next person. Do not overwrite last quarter’s scores.
The nine traits
Nine traits — score managers honestly.
Ownership. 1-5. Takes responsibility for outcomes, no blame-shifting.
Communication. 1-5. Clear, timely, two-way communication. Closes the loop.
Judgment. 1-5. Makes sound decisions without being told every step.
Emotional maturity. 1-5. Stays steady under pressure, regulates reactions.
Coachability. 1-5. Seeks and applies feedback, not defensive.
Follow-through. 1-5. Closes the loop, does what they said by when they said.
Trains others. 1-5. Can teach a task and bring others up to standard.
Handles conflict. 1-5. Addresses tension directly and fairly, does not avoid it or blow it up.
Values alignment. 1-5. Models company values when no one is watching.
Also write: Name, Role (current title), Notes (evidence, specific gaps to close, target re-eval date), Total (auto-sum of the nine, max 45), Recommendation (Promote / Develop first / Not yet).
The 1-5 anchors
1. Not yet / recurring problem.
2. Inconsistent, needs heavy supervision.
3. Developing. Does it when reminded.
4. Solid. Does it on their own most of the time.
5. Consistently strong. Others learn from how they do it.
Promote. About 37-45. Ready to lead now. Strong and even across traits.
Develop first. About 27-36. Real potential with named gaps. Give a development plan and a date. Do not promote yet.
Not yet. 26 or below. Performs tasks but is not ready to lead people. Revisit later, or keep growing them as an individual contributor.
Override rule. Any single trait scored 1-2 on Ownership, Emotional maturity, or Values alignment caps the recommendation at Develop first, regardless of total. Those are the floors for putting someone over people. Call it out when it triggers.
The bands are guides, not hard cutoffs. The owner decides. The score is an input, not a verdict. Never treat the number as a must-promote or must-pass.
How to fill it without lying to yourself
Fill without lying — anchors beat vibes.
Score behavior, not the person. Tie every number to something you actually saw. Never score personality, background, age, accent, health, family situation, or “culture fit” as a stand-in for a protected characteristic. If that is the reason in your head, redirect to what they actually did.
Watch three biases. Halo: great tech, so you assume great leader. Recency: one good or bad week coloring everything. Similarity: rating people like you higher. Name it if you see it.
If several traits are untested, say so out loud. Lower your confidence. Put a trial of responsibility in front of the decision: run a job, train a hire, own a file end-to-end. Then re-score.
After a Develop-first result, the next move is usually an accountability conversation: here is what is between you and the seat, here is the 30-day or 90-day target. After a Promote, hand them one area end-to-end and put them on the 90-day doer-to-leader spine (Weeks 7-8: develop one manager). Put every scored name on a bench list so you are not keeping five people “on your radar” and developing none.
If you want the packaged scorecard
You can run the nine traits on a legal pad. Buy Now is the Notion table delivered by email after checkout. Duplicate it so the master stays clean. The nine scores, the Total, the Recommendation, and the Notes field are already laid out. Same Square button at the top of this page.
Pairs with the Leadership Readiness Checklist (lighter yes/no read on the same person), the Restoration Leadership Bench Builder (develop the Develop-first group), the Accountability Conversation Planner (the “here is what is between you and the promotion” talk), and the 90-Day Doer-to-Leader Transition Plan. Matching Claude skill: middle-manager-scorecard. Decision support only. Not legal or HR advice. Not a hiring, firing, promotion, compensation, or disciplinary determination.
