Meta-Prompt
A prompt that instructs the model to generate or improve other prompts.
Full Definition
A meta-prompt tasks the model with prompt engineering itself — either writing a new prompt from a task description, critiquing an existing prompt, or generating a set of prompt variants for testing. This exploits the model's deep familiarity with what makes prompts work and can surface phrasing or structures the human author would not have thought of. Meta-prompting is widely used in automated prompt optimisation pipelines where prompts are evolved programmatically. It requires carefully scoping the output format so the generated prompt is directly usable without manual cleanup.
Examples
'You are an expert prompt engineer. Write a system prompt for an AI coding assistant that should help junior developers debug Python errors.'
'Critique the following prompt for ambiguity and rewrite it to be clearer: [prompt text]'
Apply this in your prompts
PromptIt automatically uses techniques like Meta-Prompt to build better prompts for you.
Related guides
Related Terms
System Prompt
The instruction layer sent before any user message that controls how an AI model…
View →Iterative Prompting
Refining a prompt through multiple rounds of feedback until the output meets req…
View →Prompt Template
A reusable prompt structure with placeholders that are filled in at runtime.…
View →