This worksheet does the work Claude used to do in the philosophy interview. You're going to answer five questions, in your own words, with the most specific and least generic language you've ever used about how you program. The probes after each question are your specificity check — if you can't answer the probe, your main answer wasn't specific enough yet. Fix it and move on.
When you're done, your completed worksheet IS your Programming Philosophy knowledge entry. You'll download it, upload to Project Knowledge, and paste the Synthesis Prompt. Claude will read your worksheet and generate one output block — your Personal Voice Rules — which you'll paste into Project Instructions. Then you're built and ready for Hot Seat #1.
What we're after: A real disagreement you have with your field. Specific enough that another clinician could argue back.
Most clinicians don't progress patients aggressively enough.
Most PTs hold patients in isometric hamstring work for three weeks before loading the hinge — by then the client has lost trust in the leg.
If you wrote a principle (a 'should' or a 'most clinicians don't'), rewrite your answer naming the actual practice, exercise, or decision you've seen go wrong.
What we're after: Your three earned defaults. Defaults you've kept because they work for your population, not because they're familiar.
Squat, hinge, push — they're good compound movements.
Hex-bar deadlift — safest posterior-chain loading for clients with limited hinge mobility. Heel-elevated goblet squat — gets depth without burning session time on ankle drills. Single-arm DB row — addresses the asymmetry I see in ~70% of desk-bound clients.
For each default, would another clinician in your discipline pick something different here? If yes, you've earned this default. If no, it might be a habit you've never examined.
What we're after: Actual numbers. Sets, reps, intensity, rest. Not "appropriate intensity" — the actual scheme you reach for first.
Moderate intensity, 3 sets of 8–12 reps, appropriate rest.
3×5 at RPE 7 for primary lifts, 3×10 at RPE 8 for accessories, 90 sec rest. For my 50+ general-health population, this hits a strength stimulus without the recovery cost of true heavy work, and the RPE 7 cap keeps form clean when I'm not in the room.
If you wrote a range, what's the single value you reach for first? Write that one. Ranges hide your decision; defaults expose it.
What we're after: The things you won't do, named, with the reason. This is the part Claude must never recommend or use.
I avoid high-risk exercises and bad communication.
No upright row (impingement risk in 50+ shoulders isn't worth the ROM). No bilateral leg press for new clients (replaces the hinge they need to learn). Never frame an exercise as 'fixing' a movement — language sets expectation, expectation sets compliance.
Did you list only exercises? Add the framings and language patterns you avoid. Words like 'compensation,' 'instability,' 'fragile,' 'wear and tear' — what's on your no-fly list?
What we're after: The orientation. The "here's how I work" that would let a sharp colleague draft something you'd be willing to sign your name to.
I care about my patients and use evidence-based practice.
I work with clients who've been told they're broken — most of my work is rebuilding their relationship with load. I move fast through exercise selection and slow through coaching. If you draft something safer than I'd write, you've drafted wrong. If you use nocebic language, you've made my work harder.
If a smart colleague read only this answer, would they know how to draft a program you'd be willing to sign your name to? If no, add what's missing.
This section feeds Claude's Personal Voice Rules generation. Be concrete. One- or two-word phrases are fine — quantity matters more than prose.