A free, open-source AI prompt that generates personalized vertical jump training programs. No one should have to pay for a vertical jump program.
Paid jump programs cost $50-$200+ and give everyone the same generic template. This prompt uses AI to analyze YOUR specific numbers and weaknesses to build a program tailored to you — for free. Just fill in your stats, paste into any AI assistant, and get a full training program with diagnostics, weekly programming, and an 8-week progression plan.
- How It Works
- Quick Start
- What the Prompt Produces
- Example Outputs
- Supported AI Models
- Contributing
- License
The prompt encodes a 6-step coaching framework that guides the AI through a structured analysis:
- Profile Analysis — Calculates diagnostic ratios from your numbers (squat-to-bodyweight ratio, clean-to-squat ratio, approach-vs-standing vertical gap) and compares them against specific benchmarks.
- Identify Limiters — Ranks your weaknesses in order: maximal strength, rate of force development/power, reactive/elastic ability, and technical/coordination.
- Build Program — Designs a weekly training schedule around your existing commitments, following programming rules for CNS management, plyometric volume, and exercise ordering.
- Provide Program — Lays out a day-by-day schedule with exercises, sets, reps, load guidance, rest periods, and coaching cues.
- Progression Plan — Outlines how the program evolves over an 8-week block with retest points and decision criteria.
- Monitoring — Lists signs the program is working, overtraining red flags, and when to move to the next phase.
The prompt isn't just "write me a jump program." It encodes specific diagnostic benchmarks (e.g., 1.5x/2.0x squat-to-BW thresholds, 60%/75% clean-to-squat ratios, approach gap ranges) so the AI has concrete criteria to evaluate your profile against.
- Open
PROMPT.md - Copy everything below the divider line
- Replace all
[BRACKETED]parameters with your information (seeexamples/INPUT.mdfor a filled-in example) - Paste into any AI assistant (Claude, ChatGPT, Gemini, etc.)
- Get your personalized program
- Diagnostic ratios calculated from your actual numbers
- Ranked list of your specific limiters with explanations
- Day-by-day weekly training schedule that fits around your commitments
- Exercise prescriptions with sets, reps, load, rest periods, and coaching cues
- 8-week progression plan with week-to-week changes
- Retest schedule and decision points for when to shift focus
- Monitoring criteria for progress, overtraining, and program transitions
These were all generated from the same input (examples/INPUT.md):
| Model | Output | Notes |
|---|---|---|
| Claude Haiku 4.5 | examples/claude-haiku-4-5.md |
More detailed analysis and coaching cues |
| ChatGPT (free) | examples/chat-gpt-free.md |
More concise, still usable |
| Claude Opus 4.6 | examples/claude-opus-4-6.md |
Most thorough analysis and programming |
| Gemini 3 Fast | examples/gemini-3-fast.md |
Colorful analogies, solid programming |
All four models produced solid programs. Claude tends to be more thorough in its diagnostic reasoning; ChatGPT is more concise and action-oriented; Gemini leans into vivid framing. All are usable.
- Tested: Claude, ChatGPT, Gemini
- Works with: Any capable LLM
The prompt is model-agnostic. It works by giving the AI a structured framework with explicit benchmarks, so any model that can follow detailed instructions will produce a useful program.
Ideas for contributions:
- Add model outputs — Run the prompt through a different model and submit the output as
examples/{model-name}.md - Improve benchmarks — Suggest better diagnostic thresholds backed by research
- Sport-specific variants — Adapt the prompt for basketball, volleyball, track & field, etc.
- Translations — Translate the prompt into other languages
To contribute:
- Fork the repo
- Make your changes
- Submit a pull request
For example outputs, use the naming convention examples/{model-name}.md (e.g., examples/gemini-pro.md).
This project is licensed under the MIT License.