Practical guides on URL slugs, AI output cleanup, JSON validation, code style, and the workflows behind the tools.
Original dataset: exact tokens per word for GPT-5, GPT-4, and Claude across seven languages, code, JSON, and CSV, with the cost of a million words by model.
AI Workflow · PillarExtractive versus abstractive summaries, how frequency-based summarization works, when each is the right tool, how to pick a length, and the privacy case for local.
AI Workflow · PillarWhat a prompt template is, the four parts of a strong prompt, why reusable templates beat one-off prompts, how to write your own, and the techniques that raise quality.
Developer · PillarWhen to convert JSON to CSV and back, the structure that converts cleanly, the Excel delimiter trap, how quoting and escaping work, and the limits of nested data.
Writing · PillarHow sentences are counted, what average sentence length says about your writing, how to fix run-on sentences, and why sentence length drives every readability score.
Writing · PillarWhat a syllable is, how to count syllables by hand and automatically, the 5-7-5 haiku rule, and how syllable count drives meter, songwriting, and readability.
AI Workflow · PillarWhat a token actually is, why GPT and Claude bill by it, and how to count tokens and estimate API cost before you send a single prompt. The browser-local way, with real tiktoken.
Writing · PillarWhat Flesch Reading Ease, Flesch-Kincaid, Gunning Fog, and SMOG actually measure, what score to aim for, and how to lower the grade level of your writing without dumbing it down.
Reference · PillarA long look at TextKit, the free, private, browser-local suite of text utilities for writers, developers, and SEOs. Thirty-three tools, forty-one long-form guides, and a single design principle: your text never leaves your machine.
SEO · PillarClean slugs lift CTR up to 39%, signal topic to Google, and now influence whether AI search engines cite your page. The seven rules. And the mistakes that quietly tank rankings.
Reference · PillarA morse code converter is a lookup table with audio attached. The 1844 origin, ITU vs American Morse, the PARIS WPM standard, audio synthesis, and where Morse still earns its keep in 2026.
Design · PillarA bold text generator does not bold anything. It swaps each letter for a Unicode character that already looks bold. The mechanism, the seven styles, and where it breaks.
Reference · PillarHow Pig Latin actually works. The two rules, the way vs yay vs hay variants, consonant clusters, the schoolyard history, and reverse-translation gotchas.
Developer · PillarHow a translator for binary turns text into base-2 and back. The byte, the bit, UTF-8 vs ASCII, emoji, worked examples, and the encoding mistakes to avoid.
Design · PillarHow invisible text actually works. The Unicode characters that look blank, where they paste cleanly, and where they get stripped.
Writing · PillarWhat character count actually measures, where it matters (Twitter 280, SMS 160, meta description 155-160, Google Ads 30), why emojis and accents break naive counts, and the workflows that depend on getting it right.
Writing · PillarA random word generator is a word list plus a coin flip. The interesting part is what you do with the output: unblock a writing session, force a brainstorm sideways, build a Mad Libs prompt, or assemble a passphrase that survives a decade of brute force.
Design · PillarWhere lorem ipsum actually comes from, why designers still use 500-year-old scrambled Latin, when it hurts the work, and the modern alternatives.
Writing · PillarThe five case conversions that cover 95% of needs: UPPERCASE, lowercase, Title Case, Sentence case, and capitalize-each-word, with the style-guide rules, the Unicode gotchas, and the spreadsheet and code patterns.
Email · ComparisonThe phrase "email extractor" hides four different products with four different price tags and four different privacy postures. The right one is rarely the most expensive one, and rarely the first result on Google.
AI Workflow · PillarEvery AI-generated document costs five minutes of cleanup before it ships. The twelve operations that matter, the workflow that compresses the tax to under a minute, and the small browser tools that do the work.
AI Workflow · How-toFive steps to take rough AI output from chat to production. Strip preamble, fix tables, wrap lists, generate slugs, dedupe. The whole pipeline is under a minute once the tools are bookmarked.
DeveloperThree places JSON validation lives in 2026. Each one wins in a different scenario. The decision tree that fits on an index card, plus the trap that costs everyone a few minutes a day.
WritingThe 2026 sweet spot is 1,500–2,500 words for most content types. But only when every word does work. The data, the breakdown by content type, and the padding test that finds filler before Google does.
DeveloperFive case conventions, five contexts. The decision framework for choosing between camelCase, snake_case, kebab-case, PascalCase, and SCREAMING_SNAKE. With code examples.
SEO · How-toSix steps to turn any title into a clean, SEO-friendly URL slug. Or one paste into a generator. Worked examples for English, accented Latin, and mixed-content titles.
AI SearchPages can rank well in Google and never get cited by Perplexity. The reason is the URL. And the asymmetry between the two systems is widening fast.
SEOHyphens are word separators. Underscores are word joiners. Google's parser has treated them differently since 2008. And the rule still matters in 2026, especially for AI search.
Markdown · PillarThe complete reference for markdown tables. Syntax, alignment, escaping, paste-from-spreadsheet workflow, and rendering differences across GitHub, Reddit, Notion, Obsidian, and Discord.
Email · PillarHow email extraction actually works. The regex, the obfuscation patterns, the deliverability check, and the legal limits.
Lists · PillarThe six list operations every workflow needs. Sort, dedupe, shuffle, prefix, suffix, reformat. With the tools and the patterns that compound the savings.
Markdown · How-toThree steps from blank document to working markdown table. Hand-typed in a minute, pasted from Excel in ten seconds. The exact keystrokes, with worked examples.
Email · How-toThree steps from raw text to a deduplicated email list. The browser shortcut, the command-line path, and the cleanup step most workflows skip.
Lists · How-toFive sort orders. Alphabetical, numeric, by length, reverse, random. And how to get each in the browser, Excel, or shell.
Lists · How-toWhy Fisher-Yates is the right shuffle algorithm, the wrong shuffles to avoid, and when a uniform random permutation actually matters.
Lists · How-toWrap every line with custom text. For HTML, markdown, JSON, SQL, or CSV. The five patterns and the cleanup step that matters.
Lists · How-toThree edge cases break most dedup workflows. Case, whitespace, order preservation. The patterns that handle all three.
Lists · How-toFrom plain list to SQL IN clause, JSON array, HTML list, or markdown bullets. With the escaping rules each destination needs.
Markdown · ReferenceEvery markdown table syntax pattern on one page. Alignment, escaping, line breaks, and the differences that matter across GitHub, Reddit, Notion, and Obsidian.
Markdown · ReferenceMarkdown tables vs HTML tables. When to use each, the trade-offs, and the workflow for embedding raw HTML inside markdown documents.
Email · ReferenceSix email regex patterns side by side. Simple, RFC-strict, obfuscation-aware, plus the language-specific syntax differences.
Email · ComparisonHunter.io vs free email extractors. What each does well, what each doesn't, and the questions that decide which to use.
Every post links to the free TextKit tools that solve the problem in seconds. No signup, all browser-local.