PPQS
WILD CORPUS · reddit

PQS 25 (F) - prompt from www.reddit.com

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

Score

F
25 / 80
gemma4:latest · local · pqs-v2.0 · canonical
Clarity7 / 10
Specificity3 / 10
Context8 / 10
Constraints2 / 10
Output format1 / 10
Role definition1 / 10
Examples2 / 10
CoT structure1 / 10

The prompt

I've been deep in prompt engineering for a while now, and one thing that always frustrated me was how scattered everything was — good prompts buried in threads, saved in random notes apps, or just forgotten.

So I put together a free library of 200+ prompts organized across 18 categories (writing, coding, SEO, marketing, MCP workflows, and more). You can browse by category or search for exactly what you need.

No signup, no paywall — just a clean page you can bookmark and actually use.

👉 [https://promptflow.digital/prompts](https://promptflow.digital/prompts)

A few of my personal favorites in there:

\- \*\*Prompt optimizer:\*\* \*"Score this prompt 1–10 on: clarity, specificity, output-readiness, and role definition. Then output a rewritten version that scores 9+ on all four."\* — I run every prompt I write through this before using it seriously.

\- \*\*Chain-of-thought injector:\*\* \*"Take this prompt and add chain-of-thought reasoning instructions so the model thinks step by step before giving the final answer."\* — Simple but it genuinely changes the output quality on complex tasks.

\- \*\*Ruthless editor system prompt:\*\* \*"You are a ruthless editor. Your job is to cut every word that doesn't earn its place. You reduce every piece of copy by at least 30% without losing meaning. You prefer short sentences. You hate adverbs."\* — Set this as your system prompt once and you'll never go back.

Would love to hear which categories feel thin or what you'd want added — I'm actively building this out and community input genuinely shapes what goes in next.

What's a prompt you keep coming back to?

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.