A local proxy that converts websites and APIs to clean Markdown. Convert HTML pages, JSON APIs, and dynamic sites. Get token counts for LLM budgeting.

Turn any website into a token-efficient JSON prompt for LLM vibe coding (ChatGPT, Gemini, Claude, DeepSeek)

A smart, LLM-powered web scraper that uses Crawl4AI to crawl any website, extracts structured data using LiteLLM, and outputs clean, tabular formats (JSON/CSV). It intelligently identifies and organizes page content into tables using AI—ideal for dynamic scraping tasks

kagiforge-libraries This is a collection of open-source libraries created by OptionFlow and used on the website https://kagiforge.com These are unique libraries for encryption, QR code generation, and converting JSON into the Toon format to reduce token usage in LLMs. An NPM package will be available soon!

A simple Python script that will parse a website and send the primary elements to an LLM to be converted into JSON

scrape data from minimalist website using Firecrawl for LLM, then using Gemini to write a python code in Colab to process the json and get the products

Open JSON standard for AI-ready metadata. LLM-Graph helps websites, documents, apps and products become understandable and actionable by Large Language Models. Includes schema, examples, plugins and validator to make any content LLM-friendly.

An automation app to discover company websites, scrape them with Playwright, extract clean content with Trafilatura, analyze them with an LLM, validate the output with Pydantic, and save results to JSON and SQLite.

NLWeb (Natural Language Web Interface) lets users interact with websites using plain language. Powered by LLMs, RAG, and JSON-LD, it enables conversational search, navigation, and information retrieval, bringing AI-driven accessibility to the web.

The `vals_parser.py` script aims to parse LLM costs/latency/benchmarks data from the website vals.ai and make it available through the charts in the README file and the 'benchmarks_data.json' file

Automated AI-powered website QA tool, uses Puppeteer to capture multi-viewport screenshots (mobile, tablet, desktop) across 3 scroll sections, then sends each image to an LLM vision model for deep UI/UX analysis. Returns a structured JSON report covering responsiveness, layout, accessibility, grammar, and UX issues.

Watched AI Labs YouTube videos about using Cursor to make better custom designs. Covered 3 methods in each video. One can screenshot websites then use LLMs to get a full JSON output breakdown to be used as Cursor input, and/ then go to the available free UI-UX component libraries to get raw code to bring to Cursor. Follow-up video went in-depth.

Versatile web scraping tool designed to extract and structure data from any website, dynamically adjusting to the data structure without hardcoded parameters. The scraper uses Selenium and BeautifulSoup for web interaction and HTML parsing, while the Groq LLM models handle data alignment and unification. The data is output in JSON