Prepare, plan, execute, review for non-code work
A designer runs brand-system work through Claude Code using plan mode, then a review prompt that lists issues only so they can choose fixes.
- Agent or tool
- Claude Code
- Topic
- Design & Creative
- Reported on
- Added
- 2026-10-15
Why it matters
Creative work still needs a human eye, but agents can speed up the drafts, variations and production steps around it. It is one of 22 sightings filed under Design & Creative on Agent Sightings (number 11 by interest), and one of 255 involving Claude Code.
How you could try it
Here is a reusable prompt, written by us in our own words, that captures the task. Paste it into Claude Code or any similar agent and adjust the details to your situation.
Review the output against the brief and the attached references. List issues only, numbered, without suggesting fixes. I'll pick which ones to address.
The prompt breaks down into 3 steps you can check as the agent works:
- Review the output against the brief and the attached references.
- List issues only, numbered, without suggesting fixes.
- I'll pick which ones to address.
Tip: Start small: run it once on a low-stakes example, read what the agent did, and only then let it run on real work.
Source
Read the original first-hand account on Reddit: https://www.reddit.com/r/ClaudeAI/comments/1n4t349/my_prepre_plan_execute_revie…. Our summary is a paraphrase; details are in the source.
Featured in these guides
Built an AI agent? Feature your agent on Agent Sightings, free submission or a featured spot.
Browse more like this
Related sightings
- Write better CSS
- Figma MCP to pixel-perfect UI
- Gamma builds the Gamma Agent for conversational editing on the Vercel AI SDK
- Wolters Kluwer drives consistent software quality at scale with agentic AI
- Book the best-priced flight and later refund it
- A contractor's personal Claude Code agent allegedly ran inside a secure Pentagon system for two months