meta-prompting

Optimize prompts through iterative scoring and store versioned templates.

Updated Apr 29, 2026

Search Tags:#optimization #llm #memory #meta-prompting #prompt-engineering #prompt

Ready to run

npx skills add https://github.com/MantisWare/BizForge --skill meta-prompting-mantisware

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: meta-prompting
Source: https://github.com/MantisWare/BizForge/tree/main/library/skills/ai-patterns/meta-prompting
Command: npx skills add https://github.com/MantisWare/BizForge --skill meta-prompting-mantisware

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Meta-prompting enables AI systems to self-improve prompts by treating the LLM as both executor and optimizer, leading to higher-quality outputs across tasks.

Core Features & Use Cases

Iterative prompt refinement: promotes ongoing improvements across prompts (task prompts, system prompts, few-shot prompts)
Structured evaluation: uses scoring and templates to guide improvements and track baseline changes
Scalable memory integration: stores optimized prompts for reuse, versioning, and retrieval in future tasks

Quick Start

Provide an initial meta-prompt optimization pass on a given prompt and save the result as the new baseline.

Frequently Asked Questions about meta-prompting

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

What is meta-prompting and how does it improve LLM outputs?▼

Meta-prompting improves LLM outputs by treating the AI as both executor and optimizer, applying meta-level reasoning to iteratively refine prompts for higher clarity, consistency, and effectiveness across tasks.

How do I automatically optimize and score repetitive task prompts?▼

You can automatically optimize task prompts by applying an iterative improvement loop that evaluates baseline prompts using structured scoring, generating refined versions and storing them as new baselines for future use.

Can I use this approach to refine system prompts and few-shot prompts?▼

Yes, meta-prompting is designed to optimize various prompt types including repetitive task prompts, system prompts, and few-shot prompts across AI workflows to consistently improve output quality.

What is the best way to manage and version optimized prompt templates?▼

The best way to manage optimized prompts is by integrating with a memory system that stores, versions, and retrieves templates, ensuring you can track baseline changes and reuse effective prompts.

Does prompt optimization work without external dependencies?▼

Yes, prompt optimization works without external dependencies because it leverages the LLM itself to evaluate and refine prompts internally, requiring no additional components to execute the improvement loop.