I built Kitzo because I was tired of searching for the same developer tools repeatedly
Hey everyone,
I recently launched Kitzo, a personal project built with Vue.js and Go:
I originally built it for my own needs.
As a developer, I regularly need small tools for one specific task. The tools usually already exist, but I have to search for a good one, try several websites, and then either bookmark it or search for it again later.
For example, I once needed a simple way to view CSV data as a readable table. Another time, I had logs containing Base64-encoded values and needed a quick decoder.
I have many examples like that: formatting JSON, decoding JWTs, comparing text, converting data, testing regular expressions, generating UUIDs, and so on.
Eventually, I decided to build Kitzo—a single place where I could keep all these tools instead of maintaining a large collection of bookmarks.
The goal is to make the tools:
- Easy to find
- Fast to use
- Consistent in design
- Available in one place
- Useful without requiring an account
It is still a growing personal project, and I would really appreciate honest feedback about the overall experience.
Does the idea make sense? Is the site easy to navigate? Would you use or bookmark something like this? What would you improve?
This resonates — building solo, I keep re-discovering the same handful of libraries every time I start something new (icon packs, error tracking, transactional email). How are you sourcing the initial tool list — scraping GitHub trending, or just your own bookmarks?
This definitely makes sense. Most devs end up rebuilding or re-finding the same small tools over and over.
The value here isn’t just the tools, it’s reducing friction — having everything fast, consistent, and in one place.
A few thoughts:
The biggest challenge will be differentiation. There are a lot of “toolbox” sites, so UX and speed matter a lot.
Search and discoverability are key — being able to instantly find the right tool without digging is what would make me come back.
“Sticky” features could help (recent tools, favorites, history).
Personally, I’d bookmark something like this if it saves me even a few minutes repeatedly.
Clean idea — execution and convenience will decide how far it goes.
There is one competitor nobody in this thread has named, and I think it is the one that actually matters: the LLM already open in the developer’s other tab. Disclosure, I work at Ojin, an AI company, which is precisely why I would take this seriously rather than wave it off.
In 2026, "format this JSON", "decode this JWT", "turn this CSV into a readable table" are all things a lot of developers now just paste into a chat window. No navigation, nothing to remember, and it tolerates fuzzy intent, which accidentally solves the recall problem raised upthread. A command palette competes with bookmarks. It does not compete with that.
That is useful though, because it tells you exactly which tools are durable. The ones that survive are where handing the data to a model is either a bad idea or a worse answer.
First, data you must not paste anywhere. Production JWTs, secrets, customer records, internal logs. A tool that runs entirely client-side and provably never transmits the input has an advantage no hosted model can match, and that is a compliance argument rather than a convenience one. If Kitzo is client-side, that should be the loudest thing on every tool page rather than a footnote.
Second, determinism. A regex tester shows you what actually matched. A model tells you what it believes matched. For anything shipping to production you need the first one.
Third, volume. Fifty thousand rows of CSV into a table view versus into a chat context window is not a close comparison.
So the sharper positioning may be "the tools you should not hand to an AI" rather than "the tools whose URL you keep forgetting." The second is a bookmark problem and it is shrinking. The first is a real and growing need, and it is defensible.
This is excellent feedback, and I agree that an LLM is now one of the main competitors for many small developer tools.
However, I don’t think it completely replaces purpose-built utilities, even for simple tasks.
For example, when I need to decode base64 or format json, I can paste the value into Kitzo and get the result in milliseconds. With an LLM, I need to write or reuse a prompt, wait 5-10 seconds, extract the result from the response, and sometimes verify that nothing was changed or omitted. That is a lot of overhead for a deterministic operation.
Fuzzy intent is extremely useful when the task itself is unclear. But for operations such as encoding, decoding, hashing, formatting, diffing, or regex testing, flexibility can become a disadvantage. I don’t want the tool to interpret what I meant—I want it to perform an exact operation and return a predictable result.
The interface also matters. Viewing 50,000 CSV rows in a searchable table, inspecting JSON as a tree, highlighting regex matches, or comparing two documents side by side is very different from receiving text inside a chat response.
And privacy is probably the strongest point. I cannot responsibly paste production JWTs, internal logs, customer records, or other sensitive data into a hosted model. For tools that execute locally in the browser, Kitzo should make that guarantee much more visible. I completely agree with you there.
So I see AI and Kitzo as complementary rather than mutually exclusive: AI is better for ambiguous tasks, explanations, and generating solutions.
Kitzo is better when the task needs to be instant, private, deterministic, repeatable, and presented through a purpose-built interface.
Your suggested positioning is valuable. “Fast, private, deterministic developer tools for data you should not paste into AI” may be a much stronger direction than simply focusing on replacing bookmarks. Thanks for the thoughtful feedback.
The collection makes sense, but I think the daily habit test is whether Kitzo removes the tab switch. A command palette or browser extension that lets me type "decode this JWT" and uses the current clipboard value would stick much better than another site I have to remember. I'd also keep the last five tools and inputs local so repeat jobs take one action. The metric that matters is probably repeat use of the same tool within 30 days, not total utilities opened.
Thank you for your feedback. I completely agree that a browser extension and a command palette would be valuable additions. The dashboard already displays your recently used tools, making it easy to access them quickly.
The interesting challenge with developer tools is that discovery is often harder than building the tool itself.
How are you thinking about making Kitzo become a daily workflow habit rather than just another collection of utilities developers visit once and forget?
My main problem was finding the right tool when I needed it. Also, many existing tools didn't work exactly the way I wanted. That's why I built Kitzo—to keep all the tools I need in one place. I'm always open to feedback and suggestions for improvements.
Appreciate the context.
The "right tool at the right moment" problem is a real friction point — it'll be interesting to see which workflows bring developers back most often.
I'm trying to be as open as possible with this project, and I truly appreciate all feedback. My goal is to build it for the developer community, so if you have any suggestions on how Kitzo could better fit your workflow, I'd love to hear them.
Appreciate that, Askar.
Would be good to continue the conversation outside the thread and hear more about how you're thinking about Kitzo's direction.
What's the best email to reach you on?
“I know it exists but can’t find it” is the real problem here. Bookmarks preserve location, but usually lose the intent and context behind saving something. I’d consider letting people search by the job they’re trying to do—“decode a token from logs”—rather than only by tool name or category. The strongest wedge may be reducing recall burden, not simply collecting more tools.
That's a brilliant idea—thank you! We already have a search feature, and it works very well, but your suggestion expands its capabilities even further. I really appreciate the feedback.
This resonates a lot - I run a site that collects a bunch of online tools, and the "I know this exists somewhere but can't find it" problem is exactly what pushed me to start it. The hard part isn't collecting tools, it's making them discoverable without turning into another cluttered directory. How are you handling search and categorization in Kitzo?
Kitzo already has a powerful search feature. If you search for terms like "json" or "curl," you'll see all the related tools. Example: https://imgur.com/a/aLjk4cs
However, if you're referring to the "I know there's a website with that tool, but I can't remember its name" problem, I think a browser extension would solve that nicely.
One thing I’ve learned is that if you find it useful and it helps you there are tons of others out there who it will help too. You’d be surprised how many people are dealing with the same things you are
The "no account, runs client-side" call is the right one — for JWT decoding and pasting internal logs especially, that's exactly the trust barrier that stops me from using random online tools, so leading with it is smart.
Yes, the idea makes sense! :) I've got the same graveyard of bookmarked one-off tool sites, and half the friction is remembering which site had the good JSON formatter or JWT decoder. Having them in one consistent UI solves the problem of finding them more than any single tool does.
One thing I'd genuinely use: fast keyboard access — a Cmd+K command palette to jump straight to a tool, or even a way to deep-link so I can bookmark the specific tool I use most.
Nice work shipping it. will bookmark and come back to it :)
“stop hunting five random sites / bookmark graveyards” is real daily friction. Vue + Go for a no-account hub of CSV / Base64 / JSON / JWT / regex tools is a clean bet: consistent UI beats another ad-heavy one-off decoder.
I’d bookmark it if the first tool is under two clicks and faster than Google → random site → cookie banner.
Which tool do you open most in a week — and is that the one Kitzo makes primary in the nav?
The bookmark graveyard problem is real, I have a folder of vaguely remembered JSON formatters and regex testers that I only find again by guessing at browser history. Your point about LLM alternatives losing on determinism and privacy is the part most 'just ask the chatbot' takes skip over. If I paste a production JWT into a chat window to decode it I have already made a decision about where that token lives now, a small deterministic tool never puts me in that position. I build a dev tool myself, and the thing that actually decides whether people come back is what your commenters flagged, it has to sit inside the workflow people already have open or it becomes one more tab they forget.