Files you are not allowed to upload
Splitting a contract PDF, pulling text out of a scanned internal document, or hashing a file containing customer data. Because the file never leaves your device, there is no export approval to obtain.
Why this project exists
When you need to merge a PDF or pull text out of a scan, most sites that come up in search upload the file to their own server. For an internal document, a contract, or anything containing personal data, that is not an option — either policy forbids it, or you simply cannot tell where the file ends up or how long it stays. This site exists to remove that problem. The tools here finish inside your browser without sending files or input anywhere. Close the page and nothing remains.
When this approach matters
When the value you paste is a credential
Inspecting a JWT, a request carrying an API key, or JSON pulled from a production database is routine work. Leaving those values on a third-party server is an incident in itself, which is why decoding and verification finish in the browser.
Networks without outbound access
On an isolated network, tools that call an external API simply do not work. These pages compute without further requests once loaded, so they keep working in that environment.
Operating rules
1. Browser-first when possible
If a task can finish in browser memory without server-side processing, that is the preferred path.
2. Practical over bloated
The site prioritizes outputs people actually copy, compare, and verify instead of loading every page with extra controls.
3. Visible feedback routes
Bugs and improvement ideas are reviewed through public channels tied to the repository and site owner.
4. Reviewable site information
About, privacy, and contact pages are kept public so visitors and reviewers can understand ownership and policies.
The project is currently maintained by taxijjang as an evolving personal toolkit.
Who builds this, and how?
This is a one-person project. With a single pair of hands, eyeballing thirty-plus tools does not scale, so nothing ships unless it passes an automated check first.
Checked in a real browser
Not static analysis: a headless Chrome renders every page and the checks read computed values off the screen. Because it sees colours and sizes after CSS actually applies, it catches rules that were written but never took effect.
Contrast and touch targets
Every text colour is composited against its real background and measured against WCAG thresholds. Buttons and inputs are checked for a 44px minimum. Light and dark themes are audited separately.
Horizontal overflow
Elements that spill outside the viewport are flagged across three widths, with deliberate horizontal scrollers exempted so the check does not cry wolf.
Proof that nothing uploads
File-handling tools are driven with a real file while network traffic is watched. The claim that processing stays in the browser is verified by behaviour, not by reading the source.
Date maths under unit test
Leap years, end-of-month clamping and daylight-saving boundaries are pinned by 49 assertions. That suite caught a real bug in the month-and-day difference between 31 January and 1 March.
Untranslated strings
Every page is scanned in English mode for leftover Korean. A missed translation blocks the release.
The source is public. You can verify the browser-only claim yourself through the repository and your browser's network tab.
Costs and advertising
Why is it free?
The tools run in your browser on a static site, so there is no server compute cost. The outgoings are a domain and static hosting, and they do not scale with visitors. There is nothing to charge for.
Will there be ads?
Enough advertising to cover running costs is planned. What will not appear: ads over the working area of a tool, ads that push the result out of view, or anything styled to look like a download button. The contact, privacy and terms pages carry no ads.
Can advertisers see my data?
No. What you type into a tool never leaves the page, so there is nothing for an ad script to read. Cookie details and how to turn off personalised ads are in the privacy notice.
Will there be paid features?
Not planned. Adding features that need a server would remove the reason this site exists, so anything a browser cannot do is left out instead.