A collection of single page web apps, mostly LLM generated. Hosted at tools.s-anand.net.

• Unicoder: Converts standard Markdown syntax into text styled with special Unicode characters to create "fancy text" for platforms without rich text support.
• Recall: Randomly recall list items from Markdown.
• Ideator: Fuse notes into a ChatGPT ideation prompt.
• Podcast: Automates the creation of a two-person podcast episode, from script generation to audio synthesis, using Large Language Models (LLMs).
• Image Generator: Chat to generate or update images.
• Postergen: Create posters and social media banners based on templates
• Daydream: Browse and rate creative ideas stored as JSONL.
• JSON to CSV: Converts JSON data into Comma Separated Values (CSV) format and can also prepare data for pasting into spreadsheet applications by converting to Tab Separated Values (TSV).
• PicBook: Generate a sequence of images from multiline captions.
• Page to MD: A browser bookmarklet that converts the content of the current web page, or a user's selection on that page, into Markdown format, copying it to the clipboard.
• GitHub Stars: Updates GitHub repository links in Markdown text, replacing link text with the repository's name, star count, and last pushed date.
• Google Tasks: Download all your Google Tasks as CSV or Markdown and delete completed tasks
• WhatsApp Scraper: A browser bookmarklet that allows users to scrape messages from an active WhatsApp Web chat session into JSON format.
• Discourse Thread Scraper: Bookmarklet that copies every post in a Discourse topic into structured JSON with replies, reactions, and metadata.
• X Thread Scraper: Bookmarklet to copy a tweet and all its replies into JSON (ads dropped), including metrics and computed buzz/keep scores.
• Gemini Conversation Scraper: Copy a Gemini conversation as Markdown with YAML frontmatter via a console script.
• Copy Links: Bookmarklet to copy all page URLs and asset links as TSV.
• Straive Intelligence: Bookmarklet to convert ChatGPT into a Straive-style user interface
• GitHub Summary: Generates a blog-post-style summary of a GitHub user's activity within a specified date range.
• Join CSV Tables: Join multiple tables on the first column.
• Google Suggest Explorer: Explore Google Search suggestions from different countries and get AI-powered humorous explanations.
• GitHub Users: A web-based tool that allows you to fetch publicly available information for specified GitHub users.
• Unicode: Helps users identify, view, and copy non-ASCII Unicode characters present in a given text, showing their hexadecimal and decimal code points.
• Hacker News Thread Markdown: Bookmarklet that copies a Hacker News post plus all nested comments as a Markdown list of lists.
• Hacker News MD: Fetches top or best stories from Hacker News, converts their linked content to Markdown, and prepends Hacker News metadata as frontmatter.
• Hacker News Links Extractor: Scrape main article links from Hacker News or HN Top Links.
• Google Fit: Allows you to view and summarize your Google Fit activity data obtained from Google Takeout.
• JSON Trim: Recursively traverses a JSON object or array and truncates string values that exceed a user-specified maximum length.
• MD to CSV: Extract the first Markdown table and download it as CSV. Useful to extract from ChatGPT.
• Excel Converter: Convert Excel data to JSONL, YAML, XML, or TOML.
• Transcribe: Provides real-time speech-to-text transcription using the browser's built-in SpeechRecognition API, capturing audio from the microphone.
• SpeakMD: Converts Markdown into conversational text suitable for audio narration. Useful to read out copied ChatGPT output.
• RevealJS: Provides a simple web interface to convert Markdown text into a Reveal.js HTML slideshow.
• WhatsApp View: Renders a JSON array of WhatsApp messages (typically from whatsappscraper) into a readable, threaded discussion format.
• WhatsApp: Generates a direct WhatsApp API link (wa.me) to start a chat with a given phone number without needing to save it to contacts.
• What Next: A flexible, offline-first single-page application for visual prioritization on a customizable grid (e.g., Eisenhower Matrix).
• Whats Near: Helps users discover nearby tourist attractions using geolocation, displaying them on a map with AI-generated descriptions and text-to-speech.
• SG Bike Park: Displays bicycle parking locations in Singapore on an interactive map, helping users find nearby bike parking facilities.
• Quotes Arena: A web application that allows users to compare short, AI-generated quotes side-by-side and vote for their preferred one, tracking AI model performance.
• LLM BoundingBox: Uploads an image and uses various Large Language Models (LLMs) via the LLM Foundry service to perform object detection, displaying results with bounding boxes and labels.

Tools from others:

• Render Markdown: Render Markdown to HTML using the GitHub Markdown API.
• Terminal to HTML: Convert terminal session text to HTML with ANSI colors.
• YouTube Thumbnails: Get YouTube video thumbnails.

1. Clone the repository: git clone https://github.com/s-anand/tools.git && cd tools
2. Serve the directory using any static file server. E.g. python -m http.server 8000 or npx serve or caddy file-server
3. Open http://localhost:[PORT] in your browser

1. Create a new directory for your tool
2. Add your tool's files (HTML, JS, CSS)
3. Update tools.json with your tool's metadata:

   {
     "tools": [
       {
         "icon": "bi-[icon-name]",
         "title": "Tool Name",
         "description": "Tool description",
         "url": "/path/to/tool/"
       }
     ]
   }

The site is automatically deployed to GitHub Pages when changes are pushed to the main branch. The deployment process:

1. Push changes to the main branch
2. GitHub Actions builds and deploys to GitHub Pages
3. The site becomes available at tools.s-anand.net

• index.html - Main landing page
• tools.js - Dynamic tool card generator
• tools.json - Tool metadata
• /[tool-name]/ - Individual tool directories, with an index.html and ESM script.js imported as <script type="module" src="script.js"></script>

MIT