ChatGPT
ChatGPT does best when you can name the shape of the answer up front: JSON, a table, a function signature, a chain of tool calls. Custom instructions and project files carry your rules from one conversation to the next, so whatever you paste into every message should live there. It handles long inputs well. It also obeys the latest instruction harder than the first, which is why the binding rules go at the end of a long prompt.
Models
ChatGPT prompts
All 3- Universal
Write a SQL query from a plain question
Produces one query against a schema you provide, and asks a question instead of guessing when the schema is ambiguous.
- Universal
Strict code review of a diff
A review prompt that reports only what would break in production, gives the smallest fix for each finding, and is allowed to say the diff is fine.
- Universal
Extract structured data as JSON
A strict extraction prompt that returns valid JSON only, with nulls for missing fields instead of invented values.
Configuration files
All 5- Universal
System prompt hardening block
A paragraph to add to any system prompt that reads user-supplied content. It cannot make injection impossible. It makes the common attempts fail…
Tested on Claude Sonnet · May 2026
- ChatGPT
ChatGPT custom instructions for technical work
A custom instructions pair that removes preamble, keeps answers checkable and stops confident guessing.
- ChatGPT
AGENTS.md for a TypeScript repo
A project file for coding agents that names the commands, the boundaries and what must never be touched automatically.
- Universal
System prompt for a support triage bot
A system prompt for classifying incoming support messages into a fixed set of routes, with an explicit escape for the unclear ones.
Guides
- Universal
Moving your rules into a config file
How to stop repeating the same instructions in every message, and what belongs in a system prompt against what belongs in the task.