JSON & XML Toolkit
The whole JSON & XML toolbox in a single extension — 50+ operations, an interactive Workbench, and 100% offline.
Stop installing five single-purpose extensions. JSON & XML Toolkit bundles the entire
data-wrangling toolbox — prettify, minify, sort, repair, flatten, query, validate, diff and
convert — for both JSON and XML, plus an interactive Workbench you won't find anywhere
else. Everything runs locally, in memory: no network requests, no telemetry, ever.
✨ Why this one?
- 🔒 100% offline & private — every operation runs locally in memory. No network calls, no telemetry; your data never leaves your machine. See Privacy.
- 🧰 One extension, 50+ operations — the full JSON and XML toolkit, instead of five single-purpose add-ons.
- ⚡ Interactive Workbench (
Ctrl/Cmd+Alt+J) — a two-pane scratchpad with live format auto-detection. Paste, transform, query, and pipe the result straight back into your editor.
- 🌳 Structure tree — an expandable outline of the active JSON document in the Explorer; click any node to jump to it. Lazy-loaded, so it stays fast on big files.
- ✅ Schema validation & semantic diff — validate against any JSON Schema with inline error markers, and diff two documents after normalizing formatting and key order so only real changes show.
- 🎯 Selection or whole file — every command transforms your selection, or the entire document when nothing is selected.
- 🩹 Lenient by default — comments, trailing commas, single quotes and unquoted keys are accepted (JSON5), so messy real-world data just works.
- 📊 Live status bar — validity, size, depth and key/element counts for the active JSON/XML file at a glance.
🚀 Getting started
- Open the Workbench — press
Ctrl/Cmd+Alt+J (or right-click → Send Selection to Toolkit Workbench) to paste, transform and pipe data back out.
- Command Palette —
Ctrl/Cmd+Shift+P → type JSON & XML Toolkit to browse every command.
- Right-click — in any JSON/XML/YAML file, use the JSON & XML Toolkit submenu for the common ones.
In-place transforms (prettify, minify, sort, escape, repair…) rewrite your selection. Conversions and queries open the result in a fresh editor, so your source stays intact.
🧩 Features
Fifty commands, one palette. Everything below works on a selection or the whole document.
JSON
| Command |
What it does |
| Prettify / Minify |
Format with configurable indent, or compact to one line |
| Sort Keys |
Recursively sort object keys (asc/desc) |
| Repair / Fix |
Parse JSON5-style input and emit clean, valid JSON |
| Escape / Unescape String |
Convert to/from a JSON string literal |
| Stringify / Parse |
Wrap a document as a string value, or unwrap one |
| Flatten / Unflatten |
Convert between nested and dot/bracket-notation form |
| Remove Null & Empty |
Strip null, "", [], {} recursively |
| Escape / Decode Unicode |
Convert non-ASCII to \uXXXX and back |
| Query (JSONPath) |
Run a JSONPath expression and collect matches |
| Validate against Schema |
Check the document against a JSON Schema, with inline error markers |
| Analyze |
Depth, key counts, type breakdown, byte size |
XML
| Command |
What it does |
| Prettify / Minify |
Format with configurable indent, or compact to one line |
| Escape / Unescape |
Convert to/from XML entities (<, &, …) |
| Query (XPath) |
Run an XPath expression and collect matching nodes |
| Analyze |
Elements, attributes, depth, and unique-tag counts |
Convert
| Command |
What it does |
| JSON → YAML / YAML → JSON |
Round-trip between JSON and YAML |
| JSON → XML / XML → JSON |
Round-trip between JSON and XML |
| JSON → CSV / CSV → JSON |
Tabular round-trip with proper quoting and type coercion |
| JSON → TypeScript |
Generate interface definitions from a sample |
| JSON → Go struct |
Generate Go structs with json: tags |
| JSON → Protobuf |
Generate proto3 message definitions |
| JSON → Python dataclass |
Generate @dataclass classes with type hints |
Inspect
| Command |
What it does |
| Diff (Semantic) |
Compare the active document with another file; both sides are normalized (sorted keys, consistent formatting) first, so only meaningful differences show |
| JSON Structure tree |
An expandable outline in the Explorer for JSON files; click a node to reveal it in the editor (lazy, so it stays fast on large files) |
| Status bar |
Live validity, size, depth and key/element counts for the active JSON/XML file |
⚙️ Settings
| Setting |
Default |
Description |
jsonXmlToolkit.indent |
2 |
Indentation for prettified output (2, 4, or tab). |
jsonXmlToolkit.sortDirection |
asc |
Direction for Sort Keys. |
jsonXmlToolkit.statusBar.enabled |
true |
Show the live status-bar info item. |
jsonXmlToolkit.repair.useJson5 |
true |
Accept JSON5 syntax when repairing. |
🔒 Privacy
Your data never leaves your machine. Every transformation, query, conversion and validation
runs entirely in-process, on your local device.
- 🔌 No network access — the extension contains no
fetch, XMLHttpRequest, WebSocket, or http/https/net/dns calls. It works fully offline (airplane mode, air-gapped machines).
- 🔒 No telemetry, no tracking — nothing about your documents or usage is collected or sent anywhere.
- 🧱 Locked-down panel — the Workbench webview runs under a strict Content-Security-Policy (
default-src 'none') that blocks any outbound connection at the browser level.
- 📄 Schema validation stays local — schemas are read from your own files and compiled synchronously; remote
$refs are never fetched (validation fails closed instead of reaching out).
The only addresses you may spot inside the bundle are identifier strings (e.g. JSON Schema
$ids and XML namespace URIs) that are never contacted.
Note: VS Code's own telemetry is separate and controlled by your telemetry.telemetryLevel
setting — this extension adds nothing to it.
🛠️ Building from source
npm install
npm run build # bundles the extension to dist/
npm test # runs the core unit tests
npm run package # produces a .vsix
Publishing: set publisher in package.json to your Marketplace publisher id, then vsce publish.
📄 License
MIT © contributors
| |
