Are you tired of explaining who you are, what you do, and how you want ChatGPT to format its responses in every single prompt thread? OpenAI's **Custom Instructions** feature solves this exact problem. By setting global preferences, you instruct ChatGPT to automatically apply context and formatting guidelines to all future conversations.
1. The Two Parts of Custom Instructions
The configuration dashboard is split into two questions that help structure the model's behavior:
- "What would you like ChatGPT to know about you to provide better responses?": This is where you specify your profession, background context, interests, tools you use, and where you are located.
- "How would you like ChatGPT to respond?": Here, you define tone (formal/casual), formatting preference (bullet points, JSON, plain text), writing length constraints, and overall boundaries.
2. Recommended Setup Template
For high-productivity use, try the following template configurations:
Part 1: Personal Background
- I am a Senior Web Developer working with React, Next.js, and TypeScript.
- I build clean, modern UI components with a strong emphasis on accessibility and performance.
- I value clean code, descriptive variables, and minimal dependencies.Part 2: Response Style
- Tone: Informative, direct, and professional. Skip introductory greetings or pleasantries.
- Code blocks: Provide full, runnable code blocks without placeholder comments or omissions.
- If an implementation has multiple choices, list the top 2 options with pros and cons before writing code.
- If unsure of context, ask clarifying questions instead of guessing.3. Disabling Custom Instructions for Specific Chats
If you need to start a casual, non-work thread (like planning a recipe or writing a joke) and do not want your professional settings to interfere, you can temporarily toggle Custom Instructions off in the settings page. This lets you quickly switch back to a default, clean-slate AI persona.