You finally have a “health check” for your AI tools.

Models update, quality shifts, styles drift… and you only notice when your content already feels wrong. This kit gives you a stable set of test prompts to re-run whenever a tool updates, so you can see:

Did the writing style change?

Are images suddenly more weird, noisy, or off-style?

Did video motion get jittery?

Is reasoning or structure worse than before?

Instead of vibes, you get repeatable prompts you can run across versions and compare.

✅ What you get

You’ll receive 3 files (pdf, csv, txt), packed with:

80 optimized test prompts designed as regression checks, not cute demos

Compatible with multiple AI tools:

ChatGPT / GPT-style LLMs

Claude

Midjourney

DALL·E

Stable Diffusion

Sora

Runway Gen-3

Pika Labs

All prompts tagged by tool and purpose so you can filter, slice, and generate your own PDFs or dashboards.

Each prompt is written to stress-test specific dimensions like:

✍️ Reasoning & structure (LLMs)

🧠 Instruction-following & formatting (JSON, tables, constraints)

🎨 Style consistency & lighting (image models)

📹 Camera motion, focus, and looping (video models like Sora, Runway, Pika)

👇 Tools covered & what the prompts test

🧾 ChatGPT / GPT-style LLMs

Step-by-step reasoning & math

Long-form article structure

Summaries, translations (IT/EN), tone changes

JSON output, markdown tables, safety/refusal behavior

🧠 Claude

Deep