Eighteen developer utilities built around one architectural decision: every parser, encoder, and converter is JavaScript running in your browser tab. The API responses you turn into TypeScript, the JWTs you decode, the email source you unravel — none of it is sent to a server, and each page keeps working offline once it has loaded. That's not a slogan; it's why the JWT and quoted-printable pages exist here at all, since the things you paste into them are usually secrets. The other design decision is honesty about limits, which most tool sites skip: the YAML converter states exactly which YAML features its parser covers and warns you about the Norway problem instead of silently mangling country: NO; the punycode converter tells you a clean homograph check is not an endorsement of a site; the flatten tool explains which key shapes can't round-trip rather than pretending everything does. Format conversion is the backbone — YAML to JSON with key order preserved, JSON to TypeScript with real optional-property detection, JSONL to arrays, XML both formatted and converted — surrounded by the decoders you need when text arrives mangled (HTML entities, quoted-printable email, percent-encoded URLs, xn-- domains), the formatters that make production SQL and minified XML readable without uploading either, the calculators for the bit-level chores (subnets with the binary shown, chmod with the special bits explained), and the generators for the tables and config files nobody writes from memory. Free, no signup, no upload.
The YAML ↔ JSON Converter goes both directions live with key order preserved and YAML's classic traps flagged as you type. The JSON to TypeScript Converter turns a sample API response into interfaces, inferring optional properties across array elements instead of guessing. The JSON Utilities page bundles the side-tasks: string escaping, JSONL ⇄ array, and flatten/unflatten for spreadsheet export. And the List to Comma-Separated Converter handles the humbler conversions between lines, CSV, and quoted lists.
The HTML Entity Encoder & Decoder unwinds ’ soup from CMS exports and feeds; the Quoted-Printable Decoder makes raw email source readable, soft breaks and subject-line encoded-words included; the URL Encoder & Decoder handles percent-encoding; the Punycode Converter translates xn-- domains and visualizes homograph lookalikes; and the JWT Decoder opens token headers and payloads locally, where tokens belong.
The tsconfig.json Generator builds a TypeScript config with every strict flag explained in plain English next to its checkbox, and the Timestamp Converter translates Unix timestamps in both directions with timezone handling done properly. The Markdown Table Generator turns a grid — or a spreadsheet paste — into clean table syntax with pipes escaped, and the chmod Calculator writes the permission octal or symbolic string with the special bits explained.
The SQL Formatter reindents queries dialect-aware — T-SQL GO batches, PostgreSQL dollar quotes, leading or trailing commas — without the query ever leaving your machine. The XML Formatter & Validator adds real line/column errors, a collapsible tree view, and JSON ↔ XML conversion. The Regex Tester matches live with a plain-English pattern explainer and a guard that keeps catastrophic backtracking from freezing the tab. The Subnet Calculator does IPv4 and IPv6 prefix math with the binary shown and a click-to-split VLSM planner, and the CRLF to LF Converter settles line-ending arguments with per-line evidence.
No. Every tool here runs as JavaScript in your browser tab — no server processing, no account, and no network request containing your data. For this category that's the core feature, not a footnote: JWTs, email source, and API responses are exactly the things that shouldn't transit a stranger's server. Open your browser's DevTools Network tab and verify, or load a page and disconnect: the tools keep working.
They're purpose-built implementations with their limits stated on each page. The YAML parser covers the subset that real configs use — mappings, sequences, block scalars, quoted strings, comments — and refuses anchors, tags, and multi-document input with a clear line-numbered error rather than guessing. The punycode converter implements RFC 3492 verified against the RFC's own test vectors. Where a browser primitive is the most correct tool (HTML entity decoding, charset conversion), the tools use it, which is how the entity decoder knows all ~2,200 HTML5 named entities.
Match the pattern of the garbage: & or ’ → HTML entities. =E2=80=99 or lines ending in = → quoted-printable. %20 and %3A → URL encoding. xn-- in a domain → punycode. Three dot-separated base64 blobs → JWT. They stack, too — HTML email is often quoted-printable with entities inside, so decode QP first, then entities.
Yes — each page is fully self-contained, parsers included. Once a tool has loaded, everything it does happens locally, so it keeps working with no connection at all.