MyText

161 browser-based text manipulation tools — format conversion, encoding, diff checking, line operations, extraction, and generation — built in an afternoon using vibe coding with Claude Code.

Full-Stack SoftwareAI Integration

What it is

M yText is a browser-based text processing platform with 161 tools organized across 13 categories: format conversion (JSON, YAML, CSV, XML, Markdown, TOML, ENV), encoding and escaping (Base64, JWT, SHA, URL, HTML entities), case conversion (camelCase, snake_case, kebab-case, Title Case, and more), counting and statistics (word frequency, keyword density, token estimation), line operations (sort, deduplicate, shuffle, number, trim), whitespace and formatting (indentation, newline conversion, zero-width character removal), find and replace (regex-powered, multi-replace), extraction and parsing (emails, URLs, hashtags, dates, numbers), detection and validation (JSON/YAML/XML/GeoJSON linting, regex testing, encoding detection), diff and comparison (side-by-side diff, Levenshtein distance, similarity scoring), and generation (UUIDs, passphrases, random data, fake user profiles). Built on React, Vite, and Tailwind CSS with Ace Editor for code editing, CryptoJS for hashing, Marked for Markdown rendering, PrismJS for syntax highlighting, and tiktoken for LLM token estimation. All processing happens client-side — zero tracking, zero logging, no data touches a server. Internationalized into eight languages including Japanese and Indonesian. Works offline after initial load.

What I did

Built 161 text manipulation tools in an afternoon using vibe coding with Claude Code — batches of 20 tools at a time, with ChatGPT assisting on prompt engineering
Engineered 28 format converters spanning JSON, YAML, CSV, XML, Markdown, TOML, ENV, and query string formats with bidirectional conversion and format-specific validation
Built a full-featured Markdown editor with split-pane live preview, syntax highlighting, file management, and HTML export using Ace Editor and Marked
Implemented cryptographic tools including JWT decoding with signature verification, SHA-1/256/512 hashing, MD5 generation, and Base64/hex encoding using CryptoJS
Built a diff checker with LCS algorithm, word-level precision, side-by-side and unified views, plus Levenshtein distance, Hamming distance, and similarity scoring tools
Integrated tiktoken for OpenAI-compatible LLM token estimation — a tool I use constantly when working with AI APIs
Designed a zero-tracking architecture with all processing client-side — only three items stored locally: theme preference, language selection, and favorited tools
Internationalized 161 tools into eight languages including Japanese and Indonesian with fully localized URLs and translated tool names