NxCreator 2.0 is the first version that feels complete end to end. The editor is faster, the runtime feedback loop is tighter, and the product no longer makes you switch tabs every two minutes to answer basic questions.
Over the past year we collected feedback from hundreds of developers building production Telegram bots. The recurring theme was not missing features — it was friction. Every save, every deploy, every time you needed to look up an API method broke the flow. Version 2.0 is our answer to that.
A rebuilt editor
We rewrote the editing surface around the workflows people actually use: iterate on handlers, inspect runtime output, format code, and move between docs without losing context. The result is lower latency and less UI noise.
The previous editor was built on a generic code editing library that was never designed for the tight bot-development loop we needed. We replaced it with a purpose-built surface that understands your project structure, knows which handlers exist, and surfaces relevant docs inline as you type.
- Faster open and save cycles — under 80ms on cold start.
- Integrated docs search for APIs, guides, and platform references.
- Smoother transitions between code, runtime output, and settings.
- Persistent panel state so your layout survives page reloads.
- Keyboard-first navigation for common editing actions.
Real-time hot-reload
One of the most requested features was the ability to see the effect of a code change without restarting the bot process. Hot-reload in 2.0 pushes handler updates to the running runtime in under a second. You can edit a callback handler, save, and immediately test it in your bot without waiting for a cold restart.
This required significant changes to how the runtime manages module boundaries. Handlers are now isolated enough to be swapped individually without tearing down the bot session or losing in-flight state. The technical tradeoff is a slightly stricter module format, but in practice most bots required no migration work.
// Before: full restart required on every change
bot.command('start', async (ctx) => {
await ctx.reply('Welcome!')
})
// After: handler updated live — no restart needed
bot.command('start', async (ctx) => {
await ctx.reply('Welcome to NxCreator 2.0!')
})AI that stays grounded in your project
The AI assistant now works with project context and docs context together. That means it can suggest handler changes, scene transitions, and MongoDB queries that are closer to the actual code you are shipping.
Previous AI integrations in developer tools tend to give generic advice because they have no grounding in your actual project. We solved this by feeding the assistant a live index of your handlers, your scene graph, and the relevant sections of the NxCreator docs before it generates a response. The result is advice that is specific enough to be useful without being a magic black box.
The product goal is simple: fewer dead ends between idea, code, and deployment.
Deployment and observability
Version 2.0 also ships improved deployment logs, structured error traces, and a live request timeline that shows message processing latency per handler. These are not glamorous features, but they are the ones that matter when something breaks at 2am and you need to understand what happened.
- Per-handler latency breakdown in the runtime panel.
- Structured error traces with file and line references.
- Deployment history with instant rollback to any previous version.
- Webhook health monitoring with automatic alerting.
What comes next
This release is the platform baseline. The next set of work is deeper project automation, richer deployment observability, and better collaboration primitives for teams managing many bots. We are also expanding the MCP integration so that external tools can read and write to your bot project without leaving their own context.
If you have been waiting for the right time to build on NxCreator — this is it. The foundation is solid, the iteration loop is fast, and the team is shipping consistently. Try it at nxcreate.com.
