Main Sections of Content in the Prompt Engineering Guide: A Complete Overview

The Prompt Engineering Guide organizes content into two complementary layers: seven top-level website sections for navigation and eight focused Markdown guides in the /guides directory for detailed technical reference.

The dair-ai/Prompt-Engineering-Guide repository serves as the definitive open-source knowledge base for prompt engineering techniques. Understanding the main sections of content in the Prompt Engineering Guide helps practitioners navigate both the live website and the underlying source files efficiently, whether they are browsing the web interface or deploying a custom instance.

Top-Level Website Sections

The live site at promptingguide.ai structures content into seven primary navigation buckets defined in the root README.md. These sections drive the user-facing information architecture.

Introduction

This foundational section covers LLM settings, basic prompting concepts, prompt elements, and general design tips. It establishes the vocabulary and principles necessary for understanding advanced techniques.

Techniques

The most comprehensive section, covering methods such as zero-shot prompting, few-shot prompting, chain-of-thought, self-consistency, knowledge prompting, prompt chaining, tree-of-thoughts, RAG, ART, APE, active-prompt, DSP, PAL, ReAct, multimodal CoT, and graph prompting.

Applications

Real-world implementation scenarios including function calling, data generation, synthetic RAG dataset creation, textbook generation, code generation, and workplace case studies.

Prompt Hub

Curated collections of ready-to-use prompts organized by task domain: classification, coding, creativity, evaluation, information extraction, image generation, mathematics, QA, reasoning, summarization, truthfulness, and adversarial testing.

Models

Model-specific guidance for ChatGPT, GPT-4, LLaMA, Mistral 7B, Mixtral, Code Llama, Gemini, Flan, OLMo, Phi-2, and general model collections.

Risks & Misuses

Safety-focused content addressing adversarial prompting, factuality concerns, and bias mitigation strategies.

Papers, Tools, Notebooks, Datasets, Additional Readings

Curated external resources including academic papers, software tools, Jupyter notebooks, training datasets, and supplementary reading materials.

Repository-Level Guide Files

Inside the /guides folder, the content splits into eight focused Markdown documents that provide deeper technical detail for local reading or custom deployments. The guides/README.md serves as the entry point for repository navigation.

Core Concept Guides

• guides/prompts-intro.md – Introduction to prompt engineering fundamentals and core concepts.
• guides/prompts-basic-usage.md – Step-by-step basics covering prompt syntax, few-shot examples, and best practices.
• guides/prompts-advanced-usage.md – Advanced techniques including chaining, tool use, and sophisticated reasoning strategies.

Specialized Topic Guides

• guides/prompts-applications.md – Real-world application scenarios and implementation patterns.
• guides/prompts-chatgpt.md – Model-specific prompting advice for ChatGPT and parameter tuning.
• guides/prompts-adversarial.md – Adversarial prompting techniques and robustness testing.
• guides/prompts-reliability.md – Strategies for reproducibility, evaluation, and error handling.
• guides/prompts-miscellaneous.md – Emerging ideas, community contributions, and future directions.

Technical Implementation: Rendering the Guides

The live site uses Next.js with Nextra to render these Markdown sources dynamically. Below is a simplified implementation pattern for loading guide content:

// pages/guides/[slug].tsx – dynamic route that renders any guide file
import { useRouter } from 'next/router';
import { getMDXComponent } from 'mdx-bundler/client';
import { getGuideContent } from '@/lib/api';   // custom helper to load raw markdown

export default function GuidePage({ source }: { source: string }) {
  const Component = getMDXComponent(source);
  return (
    <article className="prose mx-auto py-8">
      <Component />
    </article>
  );
}

// getStaticPaths – expose all guide slugs (derived from the file names)
export async function getStaticPaths() {
  const guides = await getGuideContent('list'); // returns ['prompts-intro', ...]
  return {
    paths: guides.map(slug => ({ params: { slug } })),
    fallback: false,
  };
}

// getStaticProps – fetch the raw markdown for a given slug
export async function getStaticProps({ params }) {
  const source = await getGuideContent(params.slug as string);
  return { props: { source } };
}

This implementation reads files from /guides/*.md (such as guides/prompts-intro.md) and compiles them into interactive pages, mirroring the structure defined in the root README.md.

Summary

• The Prompt Engineering Guide organizes content into seven top-level website sections (Introduction, Techniques, Applications, Prompt Hub, Models, Risks & Misuses, and Resources) defined in README.md.
• The repository contains eight focused Markdown guides in the /guides directory covering specific topics from basic usage to adversarial prompting.
• The live site renders these Markdown sources using Next.js and Nextra, with guides/README.md serving as the local entry point for developers.

Frequently Asked Questions

What are the main sections of the Prompt Engineering Guide website?

The website features seven primary sections: Introduction (fundamentals and basics), Techniques (prompting methods like chain-of-thought and RAG), Applications (real-world use cases), Prompt Hub (curated prompt collections), Models (model-specific guidance for ChatGPT, LLaMA, etc.), Risks & Misuses (safety and bias topics), and Resources (papers, tools, notebooks, and datasets).

How are the guide files organized in the repository?

The repository stores detailed content in the /guides folder as eight Markdown files: prompts-intro.md, prompts-basic-usage.md, prompts-advanced-usage.md, prompts-applications.md, prompts-chatgpt.md, prompts-adversarial.md, prompts-reliability.md, and prompts-miscellaneous.md. The guides/README.md file indexes these documents for local navigation.

What is the difference between the website sections and the repository guides?

The website sections represent the high-level navigation categories displayed on the live site, organizing content by broad topic areas. The repository guides are the underlying Markdown source files containing the detailed technical content, split into eight focused documents that map to those categories but provide deeper implementation details for developers and researchers.

Where can I find model-specific prompting advice in the guide?

Model-specific guidance is located in the Models section of the website and detailed in the guides/prompts-chatgpt.md file within the repository. This section covers prompting strategies for specific models including ChatGPT, GPT-4, LLaMA, Mistral 7B, Mixtral, Code Llama, Gemini, Flan, OLMo, and Phi-2.

{
  "mcpServers": {
    "instagit": {
      "command": "npx",
      "args": ["-y", "instagit@latest"]
    }
  }
}