Use a second, more rigorous AI model like GPT-5.2 Codex to act as a critical 'staff engineer' that reviews code written by your primary, more creative AI like Claude. This multi-model approach ensures high velocity without sacrificing code quality or accumulating technical debt.
with Claire Vo
Use your primary, 'creative' AI assistant (e.g., Claude) to write the initial code for a new feature. This is the 'vibe coding' phase focused on speed and getting a functional result.
Invoke your second, 'critical' AI assistant (e.g., GPT-5.2 Codex) and feed it the code changes (e.g., a git diff). Use a structured prompt that asks it to check for specific quality aspects.
Take a look at our current git diff and give me a report on the following: 1. Does the code accurately reflect the plan/diagram artifacts? 2. Are there any general code smells? 3. If we were to do this again and take a different approach to refactor code around it to overall improve this code base, what approach would be best?Review the detailed report from the 'reviewer' AI. Look for insights like discrepancies between the plan and implementation (visual bugs), classic code smells (missing dependencies), and strategic suggestions for refactoring.
After reviewing the feedback, simply instruct the 'reviewer' AI to implement the fixes and improvements it suggested. This closes the quality control loop and results in robust, well-structured code.
great, please make those improvements
Build a custom visual dev tool to create interactive UI mockups and flowcharts. Use these visual artifacts to guide an AI model like Claude Code to build software features with greater accuracy and less ambiguity, bridging the gap between human intuition and AI understanding.
Create a virtuous cycle with Zapier that analyzes support conversations, suggests new FAQs for a knowledge base, and uses a human-in-the-loop step to ensure quality and continuously improve an internal chatbot.
Build a Zapier workflow that automatically triggers on new calendar events, fetches customer data, uses Google's Gemini to summarize it, and delivers a complete brief to your meeting notes in Coda.
We've made ChatPRD affordable so everyone from engineers to founders to Chief Product Officers can benefit from an AI PM.