About Tools Thicket
A small collection of single-purpose web tools. Free, no sign-up, no upsell.
Who runs this site
My name is Ceniy Mei. I build and maintain Tools Thicket on my own — the code, the writing on each tool page, and the infrastructure it runs on. There is no company and no team behind it; when something here is wrong, it is my mistake to fix.
You can reach me at [email protected]. I read everything sent there.
What the site is for
Most small online utilities are one of two things: a thin wrapper around a formula, buried under interstitials and an email capture, that never tells you which rule it actually applies; or a free tier that uploads your file to someone's server and asks you to sign up before it will let you download the result.
The tools here try to be the other option. Each one does a single job, states the rule or standard it follows, and — for anything involving your own files — does the work inside your browser so nothing is uploaded at all. Where a tool does send something to a server, its page says so plainly, and so does the privacy policy.
How these tools are built
The tools on this site are produced with substantial help from AI, and I want to be straightforward about that rather than let you assume otherwise.
I wrote a multi-agent system called LLM OS that models a group of language-model agents as an operating system: a scheduler runs agent "processes" over a task graph, with shared memory and a ledger tracking every file they touch. When I approve a new tool, that system plans the work, researches the underlying standard, writes the implementation, and tests the result in a real browser. A dedicated testing stage cannot sign a tool off while the page is still throwing runtime errors.
Why build them this way
The research step is the actual reason. A percentage calculator is easy to write from memory; what takes time is checking that a receipt parser uses the right CLDR minor-unit digits for each currency, that a length converter uses the exact NIST factors, that a CSV parser follows RFC 4180 quoting, or that a contrast checker uses the correct sRGB linearisation constants. Automating the research and drafting is what lets one person ship tools that cite a standard instead of guessing at it.
What I do by hand
- Every tool is approved by me before it is built — nothing is generated and published unattended.
- I review the finished tool and its page, and I fix what the pipeline got wrong.
- The tools and the site host are open to inspection: everything runs client-side except where noted, so you can verify the privacy claims yourself in your browser's network panel.
Where this has gone wrong before
Generated copy is confident, and confident copy can describe features that were never built. In July 2026 I audited every tool page against its own source code and found several that did exactly that — a mortgage calculator whose page advertised an amortisation schedule and export buttons that did not exist, a converter offering precision modes it never had, and an invoice checker whose instructions and FAQ contradicted each other on the same page.
Those were fixed: the mortgage calculator now genuinely has the schedule, the extra-payment input and the CSV export; the false claims elsewhere were removed. I mention it because it is the honest failure mode of building this way, and because you should know what to do if you hit one. If a tool page describes something the tool does not actually do, email me and I will correct it.
Standards the tools follow
Where a tool implements a published rule, its page names the source so you can check it rather than take my word for it — NIST for length definitions, RFC 4180 for CSV quoting, RFC 6350 for vCard output, Unicode UAX #29 for word segmentation, CLDR for currency minor units, WCAG 2.2 for contrast thresholds, and ISO 4217 / ISO 8601 for currency and date formats.
What this site costs you
Nothing, and there is no account to create. The tools are free to use and I have no plans to put any of them behind a paywall or a sign-up form.
Running the site does cost money, so I may add advertising to cover it. If I do, the privacy policy will name the provider and say what it collects before the first ad appears. What will not change: the tools that process your files will keep doing it in your browser, and I will not sell your data to anyone. The site also uses Google Analytics to count visits, which is disclosed in the privacy policy along with everything else that is recorded.