All articles
Case study

I Thought I'd Had a Great 8 Weeks. The Log Analysis Disagreed — and It Was Right About All Three Things.

PA
PromptDock AIVerified creator — vouched by Prompt Dock
Jun 22, 2026 · 7 min read
promptdock.ai/blog

I'd logged eight weeks of training in a spreadsheet like a responsible adult — every set, every rep, every weight, plus little notes about how I felt. At the end of the block I leaned back and awarded myself a quiet, smug nod. Solid work. Felt stronger. Felt consistent. Then I made the mistake, or the very good decision, of running the whole thing through the Are You Actually Getting Stronger? Honest Read of Your Training Log prompt, and the smug nod did not survive contact with the data.

I used Grok 4.3 for this on purpose. I wanted the punchy, slightly blunt register, not a clinical report that would let me off the hook with hedging. I wasn't disappointed. The 'bottom line' opened with a sentence I'm still slightly annoyed by: progressing on one lift, stalled on another, and considerably less consistent than I evidently believed.

Three things I didn't want to hear

First, the squat. I'd have told you it was moving. The log told a different story: 90kg in weeks one, three, five, and seven. Flat. Five of eight weeks at the same number, dressed up in my notes as 'felt heavy today' and 'saving energy.' Second, my completion rate was 68%, not the 85% my memory had generously awarded me. Third — and this stung because it was so obvious in hindsight — there were four consecutive Fridays with no training logged. I'd excused each one individually ('busy week,' 'social thing') and never once seen them lined up as a pattern, which is exactly what they were.

The diagnosis behind the squat plateau

The genuinely useful part wasn't the bad news — it was the why. The analysis noticed that I always programmed squats as the last exercise of the session, by which point I was cooked, under-recovered, and mentally checked out. Its top recommendation was almost insultingly simple: move squats to the front of the workout, while I'm fresh. Three weeks after I did exactly that, I hit 95kg. First squat PR in two months. The plateau wasn't my legs. It was the running order, and a prompt reading a spreadsheet caught it where I never would have.

The wins it gave me felt earned, because the criticism was real

Here's a subtle thing I didn't anticipate. Because the analysis was willing to deliver genuinely bad news — the flat squat, the 68% — the 'top 3 wins' section actually meant something when I got to it. My bench was up 7.5kg over the block, and my overhead press had quietly crept up too, and the analysis called both out with the exact numbers. Normally I'd skim past praise because I assume it's padding. But an analyst that just told me, to my face, that I'd skipped a month of Fridays had clearly earned the right to also tell me what was working. The honesty cuts both ways: the criticism is trustworthy because it's specific, and the praise is trustworthy because the criticism was. I've never gotten that from a hype-y fitness app that thinks every workout is 'crushing it.'

Why Grok, and why this one's worth paying for

I picked Grok 4.3 for this deliberately, and the register is the reason. I did not want a soft, hedging, on-the-other-hand report that would let me wriggle off the hook. Grok delivers the read with a bit of edge — the punchy, slightly blunt coach voice that says 'your squat is flat, here's the number, fix the running order' instead of 'you might consider exploring whether your squat progression has plateaued.' For a tool whose entire job is to puncture my self-flattery, a model that's comfortable being direct is a feature, not a personality quirk.

And this is the one prompt in my fitness toolkit I'd genuinely pay for, and I have, because it does the thing I am constitutionally incapable of doing for myself: it reads my effort without my ego attached. I am a deeply biased narrator of my own training. The log isn't. Feeding it to an analyst that cites the numbers back at me, refuses to pad bad news, and hands me a prioritised change list is worth more than any single workout. I run it at the end of every block now, the way you'd reconcile a bank statement instead of just vibing about your finances.

I keep my log in the same place I build my plans with the Weekly Training Block From Your Goal, Gear, and Real Schedule prompt, so the loop closes cleanly: build the block, run it, log it, then get the honest read before I touch anything. Grab the Are You Actually Getting Stronger? Honest Read of Your Training Log prompt on Prompt Dock, paste in your last two months, and brace yourself a little. The truth about your training is sitting in your own spreadsheet. You just need something willing to say it out loud.

The prompt behind this post
Premium
Are You Actually Getting Stronger? Honest Read of Your Training Log

Paste 4-8 weeks of training logs and get a blunt, data-cited analysis: what improved, what plateaued, your real session-completion rate, your top wins and top issues, and the specific changes to make in your next block.

View promptGrok 4.3
Keep reading
Our Support Bot Confidently Made Up Enterprise Pricing — One System Prompt Fixed It

The agent wasn't hallucinating wildly, it was extrapolating plausibly. That's worse. Here's the exact system prompt that taught it to know its own limits.

Elite Prompting: Why Better Prompts Beat Random AI Instructions
I Designed Branded Wrapping Paper for My Candle Business for $28, Not $400

Packaging is part of my product, but a custom print run starts around $400. Print-on-demand plus one seamless-tile prompt got me there for the price of lunch.

Related prompts
Start in two minutes

Find a verified prompt for the job.