PPQS
WILD CORPUS · reddit

PQS 37 (D) - prompt from www.reddit.com

Source: www.reddit.com · Scraped 2026-05-04 · Scored 2026-05-04

Score

D
37 / 80
gemma4:latest · local · pqs-v2.0 · canonical
Clarity7 / 10
Specificity6 / 10
Context7 / 10
Constraints3 / 10
Output format4 / 10
Role definition3 / 10
Examples5 / 10
CoT structure2 / 10

The prompt

This Is the Exact Prompt I Use to Get 10x Better Answers from ChatGPT

# 🎯 The Prompt:

>"Act as a \[ROLE\]. I need you to \[TASK\] for \[AUDIENCE\] with \[STYLE/FORMAT\]. Include \[SPECIFIC ELEMENTS\]. Start with a summary and finish with a recommendation." 

# 

# 📌 Example:

Let’s say I want help writing a blog post.

I use:

>

The difference is wild. ChatGPT stops giving generic fluff and starts acting like a specialist.

# 🔍 Why This Works:

* You give **context** (who GPT is supposed to be)
* You define the **task** clearly
* You mention the **audience** and **tone**
* You guide the **output format**
* You add structure = less randomness

This prompt was scraped from a public source. The score reflects the input as written, not the quality of any output it produced. The AI input quality problem is the gap between what people type and what the model can act on.