Apr 18 • General Discussion 💬
Top prompting techniques
Hey !
I came across a post about advanced prompting techniques and thought it would be cool to compare notes with you all
1. Chain of Thought (CoT)
Ask the model to “think step-by-step” — perfect for complex reasoning or breaking down tasks.
2. Role-Based Prompting
Tell the model who it is (e.g. “You’re a brand strategist”) and you’ll get much more focused answers.
3. Prompt Chaining
Link multiple prompts together, feeding the output of one into the next. Super powerful for workflows.
4. Self-Reflection & Refinement
Ask it to review its own response and improve it. You’d be surprised how often it can level itself up.
5. Goal-Driven Prompting
Set a clear objective (like “convince”, “summarize”, or “inspire”) and guide the model toward it.
6. Retrieval-Augmented Prompting (RAG)
Give the model access to your own docs or data to make its answers smarter and more grounded.
7. Structured Output Prompting
Ask for answers in formats like tables, bullet points, or JSON – super useful for apps, automation, or clean notes.
Drop your favorite prompt tricks below!
Let’s gooo — one prompt at a time.
13
10 comments
Top prompting techniques
skool.com/ai-automation-society
A community built to master no-code AI automations. Join to learn, discuss, and build the systems that will shape the future of work.
Leaderboard (30-day)
1
🔥
+2468
2
🔥
+1105
3
+1027
4
🔥
+825
5
+785
Powered by