The Weekly Check-In Triaged by 3 Signals

The Weekly Check-In Triaged by 3 Signals

Decision quality beats vibe checks when the week gives you appetite, glucose, and adherence data

Justin Harris cut through a real-world retatrutide check-in with a blunt mechanism: appetite suppression, fatigue, and the risk of pushing food intake too low. The concrete finding was simple — 2 mg on Friday produced “no appetite whatsoever” through Saturday and Sunday, with more fatigue than normal — and the coaching response was equally concrete: run with it only while leaning out, then pause or reduce it if the next phase requires more food. That is the right frame for weekly check-in triage: the best decision is not the most interesting intervention, but the one that matches the current bottleneck. If the bottleneck is appetite control, suppress it; if the bottleneck is growth and food throughput, don’t let a tool that blunts appetite run the show.

That’s the thesis: weekly check-ins should be triaged by bottleneck, not by novelty.

The job of a check-in is not to gather everything

A lot of coaches collect data the way people collect screenshots: lots of it, little structure, and too much confidence that more information equals better decisions. The better model is narrower. In the examples available here, the useful weekly check-in information clusters into three decision signals:

1. Appetite and food throughput
2. Glycemic response and meal timing
3. Adherence and execution quality

Those aren’t abstract categories. They are the exact places where the coaching decisions change.

Rory Lazowski’s report on retatrutide is a clean example of signal one. After 2 mg, appetite fell hard enough that he had “no appetite whatsoever,” even on low-carb days, and fatigue increased. Justin’s response was not to celebrate the new tool as automatically “better.” He noted the appetite effect, acknowledged the first-hand test for client understanding, and then placed the decision in context: if the goal is to add food, the dose may need to be paused or reduced; if the goal is to lean out while it’s easy, keep it moving. That’s triage. The same effect can be helpful or harmful depending on the phase.

The practical lesson for coaches is not “retatrutide works.” It is: if appetite is collapsing in a phase where food must rise, that is a warning flag, not a success metric.

Glucose data only matters when it changes the next meal

The second signal is more valuable because it is more operational. Joe Webb’s check-in shows how a small change in insulin sensitivity showed up in the day’s execution: the same insulin dose as the prior week caused a noticeable blood sugar dip, not a true hypo, but enough that meal 2 had to come 30 minutes sooner. He reduced the dose by 1 IU and still had to pull the next meal forward, then planned to reduce further on the next high day.

That is the kind of data that actually deserves attention in a weekly review: not a lab result floating in isolation, but a result that altered the timing of the next meal. It is decision-relevant because it affects the whole chain of the day. Once the meal window shifts, the rest of the schedule can shift, and the supposed “plan” becomes a moving target unless the coach adjusts it.

The causal order matters here. First, insulin sensitivity improved. Second, the same dose caused a lower-than-expected glucose response. Third, meal timing had to tighten. Fourth, dose was reduced further. That is a tidy example of why check-ins should be triaged around execution breaks, not just averages.

A weak coach reads “blood sugar dipped” and starts narrating. A stronger coach asks only two questions: did it change the meal schedule, and will it happen again if nothing changes? In Joe’s case, the answer to both was yes.

The best check-in decisions are phase-dependent

The retatrutide example and the insulin example point in opposite directions, which is exactly why weekly triage is useful. One check-in says, “food intake is being pushed down too far for the next phase.” The other says, “glycemic control is improved enough that the current dose is now too much for the same meal timing.” In both cases the response is not ideological. It is phase-aware.

Justin’s own coaching language across the source set is consistent with that. He repeatedly treats tradeoffs as real, not theoretical. Appetite suppression can be a tool when the objective is leaning out. It becomes a liability when the objective is pushing calories. Improved sensitivity can be a win, but if it forces meals closer together, the current setup is no longer stable. That’s the difference between having data and having a decision.

This is where AI coaching can be genuinely useful — but only if it behaves like a triage layer, not a content generator. A model should not merely summarize that “the client is doing well.” It should answer: which bottleneck moved, what did it break, and what needs to change first?

A useful weekly check-in has three branches

For coaches, the decision tree can stay simple:

1) If appetite is the bottleneck, track intake feasibility

If hunger is high and fullness is low, then appetite control may be valuable. If appetite is already low and the next phase requires more food, that same tool may need to come off or be reduced. The retatrutide case is useful precisely because it shows the tool working too well for the wrong phase.

2) If glucose is the bottleneck, track meal timing before anything else

The Joe Webb example shows that a dose change is justified not just by the number on a meter, but by the way it forces the actual feeding schedule. When meal 2 has to move 30 minutes earlier, the coach has enough evidence to treat the current dose as unstable for that day’s structure.

3) If adherence is the bottleneck, fix the simplest break first

The nutrition material in the KB keeps pointing back to the same principle: most results come from nailing macros and execution, while the last few percent lives in the details. That matters because coaches often over-treat the details before checking whether the basics held. If the client missed the real target, no amount of sophistication rescues the week.

That last point is the quietest and most important one. A weekly check-in should not turn every issue into a pharmacology problem, a supplement problem, or a “programming” problem. Often the first question is much simpler: did the client actually execute the plan as written?

What AI should do in triage mode

If you want AI to be worth using in weekly check-in review, make it classify, not embellish.

A useful system should surface:

• What changed since last week
• Whether the change improved or worsened the current bottleneck
• Whether the bottleneck is appetite, glucose, adherence, or load management
• Which single adjustment is most likely to restore stability

That’s a better use of AI than trying to sound insightful.

The reason is straightforward: weekly coaching is mostly about stopping the wrong problem from taking over the rest of the week. Retatrutide can suppress appetite hard enough to be useful or too much to be useful. Improved insulin sensitivity can be a win or a scheduling headache. The coach’s job is to identify which side of that line the athlete is on right now.

The falsifiable test

Here is the thesis you can actually test over a month of check-ins: if your weekly review produces better decisions, you should see fewer “mystery” problems and faster corrections in appetite, glucose timing, and adherence. If the review is just collecting more data without changing the next meal, the next dose, or the next behavior, it is not triage. It is theater.

That is the standard. Weekly check-ins should not impress you. They should make the next decision easier.

Sources Used

• raw/_consumed/2026-05-26/troponiniq_kb.md
• raw/_consumed/2026-05-31/kahunas-export/2026-05-31-w13-18m/clients/rory_lazowski___members-c5balaovjbdoeefqmfuqdhh2tbpmfdu16lnf0tnrtmw.json
• raw/_consumed/2026-05-31/kahunas-export/2026-05-31-w13-18m/transcripts/rory_lazowski___members-c5balaovjbdoeefqmfuqdhh2tbpmfdu16lnf0tnrtmw.md
• raw/_consumed/2026-05-31/kahunas-export/2026-05-31-w13-18m/clients/joe_webb___members-rksigkykimaxwmo_t4_e8nwvbtc2j0etleutkyysads.json
• modules/03-knowledge/kahunas-coaching-deep-nutrition.md