AI-Friendly Documentation Project

This repository contains AI-optimized documentation for various open source projects. The goal is to provide documentation in a format that is easily consumable by Large Language Models (LLMs).

Using the Generated Documentation

The AI-friendly documentation can be used in several ways:

1. Direct Consumption by LLMs: The markdown files can be directly fed to LLMs to improve their understanding of Angular concepts

2. RAG Integration: The vector embeddings can be used in RAG systems to enhance LLM responses with accurate Angular documentation

3. Training Data: The processed documentation can serve as high-quality training data for future AI models

For more insights on improving LLM responses with AI-friendly documentation, see my article: "Getting Better LLM Responses Using AI-Friendly Documentation"

Angular 19.2.3

The Angular documentation is derived from the official source at Angular GitHub Repository.

• Complete Angular documentation in a single file, table of contents
• Individual documentation files for specific Angular features

Scripts

The angular.ts script is responsible for generating AI-friendly Angular documentation from the official Angular source files. It performs three main operations:

1. Repository Cloning: Automatically clones the specific version of the Angular repository from GitHub into a local directory.

2. Documentation Generation: Transforms the original Angular documentation into formats optimized for AI consumption.

3. Vector Embedding Creation: Creates vector embeddings of the transformed documentation using HuggingFace's embedding models. These embeddings are stored in a local Qdrant vector database (running at localhost:6333), making them readily available for semantic search and retrieval.

Output Files

The script generates:

• A comprehensive markdown file containing all Angular documentation
• Individual markdown files for specific Angular features in the sections directory
• A table of contents file
• Vector embeddings for search and retrieval

Vector Storage

The vector embeddings are stored in a Qdrant vector database with the following setup:

• Server: Local Qdrant instance running at http://localhost:6333
• Collection Naming: Each documentation set gets its own collection (e.g., angular-19.2.3)
• Model: Uses Xenova/all-MiniLM-L6-v2 for generating embeddings
• Chunking: Documentation is intelligently chunked using a LlamaIndex.TS's Markdown-aware parser that preserves document structure

Directory Structure

The script maintains the following directory structure:

./
├── cloned-repos/               # Repository storage location
│   └── angular-{version}/      # Cloned Angular repository
├── ai-friendly-docs/           # Generated documentation output
│   └── angular-{version}/      # Version-specific documentation
│       ├── angular-full.md     # Complete documentation in one file
│       ├── toc.md              # Table of contents
│       └── sections/           # Individual documentation files

To use the generated embeddings in your own applications:

1. Ensure Qdrant is running locally: docker run -p 6333:6333 qdrant/qdrant
2. Connect to the appropriate collection name corresponding to the documentation version
3. Perform semantic searches against the indexed content

// Example code for querying the vector store
import { QdrantVectorStore } from "@llamaindex/qdrant";
import { VectorStoreIndex } from "llamaindex";

// Connect to the existing collection
const vectorStore = new QdrantVectorStore({
  url: "http://localhost:6333",
  collectionName: "angular-19.2.3",
});

// Create an index from the existing collection
const index = await VectorStoreIndex.fromVectorStore(vectorStore);
const retriever = index.asRetriever({
  similarityTopK: 3,
});

// Query the documentation
const nodesWithScore = await retriever.retrieve({
  query: "Are standalone components default in Angular?",
});

console.log(nodesWithScore);

To run the script:

pnpm run generate:angular

Future Additions

More open source projects will be added to this collection in the future.