Browser-native workspace
Your data never leaves this page.
Professional tools for JSON, text, and AI — right in your browser. No uploads, no sign-up.
Proof, not promises
Don't take our word for it.
These are real tools running in this page. Watch the network while you type.
Input — paste JSON here
Result
Input — paste Base64 here
Decoded
Input — Markdown
HTML output
Input — paste any text
Stats
No servers needed. Not a trick — an architecture.
How can tools run without uploading anything? The tools download once as static files; your data is processed by your own CPU; no request ever carries your data.
-
1
The tool arrives as code, once.
HTML, JS, and WASM — fetched like any page, cached like any page.
-
2
Your browser runs it — your CPU, your memory.
Formatting, decoding, counting: this class of work needs no data center.
-
3
Your data has no route out.
There's nothing to send to. The Network tab stays quiet because the architecture is.
Verify everything
Check us. Please.
Verify it yourself — 60 seconds, no faith required. The claim on this page is falsifiable by design; here is the test.
-
Step 1 of 4
Press F12 (or ⌘⌥I) → Network tab
Every browser ships an unfalsifiable request log. Open it.
-
Step 2 of 4
Check "Preserve log", then clear
Start from an empty record so nothing hides in scrollback.
-
Step 3 of 4
Come back and run any tool
Paste something real. Format it, decode it, count it.
-
Step 4 of 4
Watch nothing appear
That absence is the product. The meter on this page reads the same log.
Security posture
No accounts — there is nothing of yours on our side to breach.▾
Check it yourself: use any tool signed out. There is no login route, no session cookie for you to find.
Anything remembered is remembered by your browser, not by us.▾
Check it yourself: DevTools → Application → Local Storage. Keys are named in plain sight (ts-*) — inspect, export, or delete them.
This site is static files — there is no application server holding your data.▾
Check it yourself: DevTools → Network. Page loads fetch static assets; nothing you type is ever POSTed anywhere.
Dependencies run client-side only.▾
Check it yourself: the Network tab lists every file this page loads — static assets, no third-party data endpoints.
The long version lives at /privacy — same claims, same tests.
Three workspaces
Three workspaces. One rule: your data stays yours.
JSON, encoding, regex, CSV — the bench you already know.
JSON · Markdown · CSV · Utilities
Enter the Developer workspace →
Prompts in, clean output back. No model ever reads them.
Prompt · Utilities
Enter the AI workspace →
Count it, sort it, compare it, fix it.
Text · Analysis
Enter the Text workspace →
Where to start
Start with these.
Developer Essentials9 tools
View all →The nine tools developers reach for daily.
JSON Beautifier
Format and indent JSON with syntax highlighting.
JSON Minifier
Strip whitespace from JSON for production use.
Base64 Decode
Decode Base64 strings back to plain text.
JWT Decoder
Inspect JWT header, payload, and signature.
Regex Tester
Test and debug regular expressions live.
CSV to JSON
Convert CSV rows to JSON objects instantly.
Markdown to HTML
Render Markdown as clean HTML.
JSON Repair
Fix broken or near-valid JSON automatically.
Cron Parser
Validate cron expressions and preview next runs.
AI Essentials5 tools
View all →Prompts in, clean output back. No model ever reads them.
Prompt Cleaner
Strip noise from AI prompts for cleaner output.
Prompt Formatter
Structure and reformat prompt text.
Prompt Token Estimator
Estimate token usage before sending to a model.
AI Output Cleaner
Remove markdown artefacts from AI responses.
Prompt Template Builder
Compose reusable, structured prompt templates.
Daily Toolkit6 tools
View all →Count it, sort it, compare it, fix it.
Word Counter
Count words, characters, sentences, paragraphs.
Case Converter
Convert text to upper, lower, title, camel case.
Line Sorter
Sort lines alphabetically or reverse.
Duplicate Remover
Remove duplicate lines from text.
Reading Time
Know what you're asking of your reader.
HTML to Markdown
Rescue clean text from pasted page soup.
Editor's picks our opinion, rotated by hand
JSON Repair
json-repair — for when the API answered almost-JSON.
CSV to Markdown Table
csv-to-markdown-table — the fastest route from spreadsheet to pull request.
Reading Time Calculator
reading-time-calculator — know what you're asking of your reader.
HTML to Markdown
html-to-markdown — rescue clean text from pasted page soup.
Cron Parser
cron-parser — parse Unix, Quartz, and seconds-prefix schedules with next-run preview.
The honest comparison
The difference isn't the tools. It's where they run.
| Dimension | Traditional upload-based tools | Browser-native (this site) | Verify it yourself |
|---|---|---|---|
| Where your data goes | Uploaded to the operator's server | Stays in this tab's memory | DevTools → Network while using the tool |
| Where processing happens | Their machines | Your machine (JS/WASM) | DevTools → Performance: local CPU activity, no request/response gap |
| Waiting | Upload + queue + download round-trip | Keystroke-speed; no transfer to wait for | Type in the live demo; output is same-frame |
| Accounts & limits | Sign-up walls, daily quotas, file-size tiers | None — the tools are static files | Use any tool signed out; there is nothing to sign in to |
| What the operator can see | Your file contents, filenames, IP + content pairing | We can see standard web-server logs for page loads — never your data, which produces no requests | Network tab: page assets vs. your activity, itemized |
| Works offline after load | No — every operation needs their server | Mostly — a loaded tool keeps working offline | Load a tool, set DevTools → Network to "Offline", keep using it |
Where your data goes
Traditional
Uploaded to the operator's server
Browser-native
Stays in this tab's memory
Accounts & limits
Traditional
Sign-up walls, daily quotas, file-size tiers
Browser-native
None — the tools are static files
Where processing happens
Traditional
Their machines
Browser-native
Your machine (JS/WASM)
Works offline after load
Traditional
No — every operation needs their server
Browser-native
Mostly — a loaded tool keeps working offline
Honest caveat: offline works after a tool has loaded; a tool you haven't visited yet still needs its first fetch. And we see page-load logs like any website — never your data.
Go deeper
Know the format, not just the button.
Learning Center
Format internals, edge cases, and which tool catches what — written by the people who build the parsers. The guides publish when they're worth your time, not before.
Publishing soon
Workflow Library
Tools are steps. These are the jobs — API response → clean JSON → CSV → Markdown table, chained on one page. The library opens with the first documented workflows.
Publishing soon
Questions
Still unsure? Fair.
Yes. Static hosting is cheap and your data never touches our costs. There are no tiers, no quotas, and no paid unlock behind any tool.
Nowhere. Tools download once as static files and run on your CPU; what you paste stays in this tab's memory. Open DevTools → Network and watch: your typing produces zero requests. Verify it yourself →
No. There is nothing to sign in to, and nothing of yours on our side to breach. Anything a tool remembers lives in your browser's localStorage — inspect it, export it, delete it.
Mostly. A tool you've opened keeps working with the network off. Honest caveat: a tool you haven't visited yet still needs its first download — full offline packaging is on the roadmap.
Only your machine's. Processing is local, so the practical limit is your browser's memory, not an upload cap. Large inputs run in a worker so the page stays responsive.
There isn't one to hide: static hosting is cheap, your data never touches our costs. If we ever add paid features, the free tools stay free — and stay local.
More questions → about the shop