Back to guide/Education & Learning

Iterative Refinement

Don't try to get perfect output in one shot. Write a draft (even a bad one), then ask the AI to improve specific aspects. This gives you more control than generating from scratch and teaches you what instructions produce the best edits.

prompt-engineering-beginnersdraftimprovement_1improvement_2

Edit View

Prompt

Here is my first draft:

"""
{{draft}}
"""

Please improve this by:
1. {{improvement_1}} (e.g., "Make the opening hook more compelling")
2. {{improvement_2}} (e.g., "Simplify the technical jargon for a general audience")
3. {{improvement_3}} (e.g., "Add a stronger call to action at the end")

Keep everything else the same — only change what I asked for. Show the full revised version.

Variables to customize

Why this prompt works

Save this prompt to your library

Organize, version, and access your best prompts across ChatGPT, Claude, and Cursor.

Related prompts

Sentiment Analysis

Requesting confidence and key phrases forces the model to justify its classification rather than guessing. The structured output format works zero-shot because sentiment analysis is well-understood by LLMs.

Key Information Extraction

Listing the exact fields to extract removes guesswork. The 'Not specified' instruction prevents hallucination when information is missing -- a common failure mode without this guardrail.

Question Answering with Source

Grounding the answer in source material and instructing the model to refuse when information is missing dramatically reduces hallucination -- the biggest risk in zero-shot Q&A.

Math Word Problem Reasoning

Explicit numbered steps force the model to decompose the problem rather than guessing. The verification step catches arithmetic errors before the final answer.