Download. Load. Run.
Not a hosted app. Not a subscription. Just files that configure your AI to work systematically.
See it in action
Here's what a kit looks like and how it transforms your AI into a systematic operator.
What you download
The system prompt
1# Discovery Machine - System Prompt2# Load this into Claude, ChatGPT, or any LLM34role: product_validator5version: 2.1.067behavior:8 - Ask clarifying questions before validating9 - Use scoring rubrics for all assessments10 - Stop when kill criteria are met11 - Output structured candidate frames1213scoring:14 feasibility:15 technical: 1-516 market: 1-517 timing: 1-518 threshold: 12 # Minimum score to proceed1920kill_criteria:21 - No clear problem statement22 - Market size < $10M23 - 3+ direct competitors with >$50M funding24 - Technical complexity > 18 months MVPWhat happens when you use it
The process
Download
Get the complete system
You receive a folder with everything: system prompts, templates, scoring rubrics, and a runbook. Plain files you can open anywhere.
Load
Copy-paste into your AI
Load the system prompt into ChatGPT, Claude, or any LLM. It configures the AI to follow the methodology—asking the right questions, using the rubrics.
Run
Follow the runbook
Work through the process step by step. Fill templates, score candidates, make evidence-based decisions. The system guides you.
Capture
Save your outputs
Save validated candidates, experiment results, and decisions. Reuse the system next time. Share with your team.
Why this works
Systems > Prompts
A prompt is a one-shot instruction. A system is templates, rubrics, and guardrails that produce consistent results.
Evidence > Vibes
Every kit has numeric scoring and kill criteria. You know when to continue and when to stop. No guessing.
Files > Apps
You own the files. They work offline. No subscription, no vendor lock-in, no "free tier limits."