Kiro: A new agentic IDE

Here my problem with this: I don't want to be jumping an editor/IDE every 6 months, learning new key bindings and even more importantly, getting used to a completely new look.

In a space that moves as quickly as "AI" does, it is inevitable that a better and cheaper solution will pop up at some point. We kinda already see it with Cursor and Windsurf. I guess Claude Code is all the rage now and I personally think CLI/TUI is the way to go for anyone that has a similar view.

That said, I'm sure there's a very big user base (probably bigger than terminal group) that will enjoy using this and other GUI apps.

The nice thing about CLI/TUI is that you can keep using your editor or IDE of choice and chat with the AI on the side.

Totally agreed, which is why I'm sticking with my editor (neovim) regardless of whatever AI thing is hot and using tools outside/analogous to it, currently claude code

> Here my problem with this: I don't want to be jumping an editor/IDE every 6 months, learning new key bindings and even more importantly, getting used to a completely new look.

You're basically advocating for GNU Emacs: https://github.com/karthink/gptel

Every 6 months? It's turning into every two weeks. Sticking with claude code. Its working beautifully for us.

They're all based vscode, so the switching costs are fairly minimal? It'll get worse over time as they diverge, but at the moment they're all fairly similar AFAICT. It's starting to become noticeable that Cursor isn't picking up VSCode enhancements and fixes, but it's still quite minor.

I have a question. I do not like the concept of "agent mode" for AI. I'm a control freak and I want to control every line that gets committed because I am responsible for it and need to understand/visualize/memorize every part of codebases I work on.

Is Claude Code good for the "ask" flow? No, right?

The old flow before agent mode got added. Select some code, ask questions about it or give an instruction on editing it and then choose to accept the change.

As I understand (I could be wrong), with agent mode, it edits the file for you, no way for you to accept before it does, so you have to manually check the diff, roll back parts you don't want, etc.

Am I right?

Thanks for the link. I'm not an emacs user and I'm more in the search of something like opencode [1], but I think it's not polished enough yet. I actually want to contribute to open source, so maybe I should create my own thing, heh.

[1]: https://github.com/sst/opencode

I like aider's solution of encapsulating each edit in a git commit; I hope that gets widely adopted.

Claude Code is definitely more agentic, but you can use it in a variety of ways. In Plan Mode it won't touch the code, and by default it asks you to accept every single diff. With IDE integration you can definitely just highlight some code and ask questions, I don't see why that workflow wouldn't work.

CLI -> Voice & Gesture UI

Not really, even at work I got to test couple different AI solutions and the experience is always slightly different, even if the editor is the same, for the most part. It's the tiny things like using the prompt template, or opening the panel. (I could, of course, make an attempt to customize the keybindings, but why bother when it changes so quickly.)

only if you use vscode, I think TUIs are a better option since a lot of us use other ides than vscode

Add two lines to CLAUDE.md and claude code can do this as well :)

It could be better. I think the PMs and investors and decision makers at these companies are running with a "we want to replace / automate developers" philosophy, while these tools are actually best at augmenting developers. And so they're sorta builing with this "I'll do everything and ask you for confirmation" (and basically encourage you to give me blanket permission).

In reality these tools would be best if they took a more socratic method, a more interactive pair programming approach. So instead of giving you a blanket diff to accept or refuse or "No, and here's some changes" -- it should be more dialog oriented.

Of all of them so far though, I think Claude Code is closest to this. IF you prompt it right you can have a much more interactive workflow, and I find that most productive.

> As I understand (I could be wrong), with agent mode, it edits the file for you, no way for you to accept before it does, so you have to manually check the diff, roll back parts you don't want, etc.

You’re sort of technically correct but I wouldn’t really describe it this way exactly. You have to proactively accept or reject all changes to your files in some way.

It is almost impossible to accidentally commit code you don’t want.

It’s not really an edit in the same sense as an unstated change. It doesn’t even really do that until you accept the result.

It’s basically saving you a UI step compared to ask mode with basically no downside.

YUP! This is why I've settled on Aider and it's "IDE integration" (watches all files for comments that end in "AI!", which then invokes the AI). I can then use it with whatever editor I prefer. I view the vscode mono-culture as a bad thing. I also like I can use any AI backend I like, which is really how it should be: Vendor lock-in tools are bad, remember?

I guess you lose tab-completion suggestions, but I am not a fan of those compared to 'normal' tab-complete (if backed by an lang server). If I want AI, I'll write a short comment and invoke the tool explicitly.

EDIT: Of course, it really depends an your usecase. I maintain/upgrade C code libs and utils; I really cannot speak to what works best for your env! Webdev is truly a different world.

EDIT2: Can't leave this alone for some reason, the backend thing is a big deal. Switching between Claude/Gemini/Deekseek and even rando models like Qwen or Kimi is awesome, they can fill in each other's holes or unblock a model which is 'stuck'.

Unfortunately aider is showing its age. It is great for what it does, but better LLMs + "agentic" have shown that you can get more in the swe domain.

There was a paper recently where they had an LLM evolve tool harnesses and got ~20% more than w/ aider on the benchmark they used, so it's pretty clear that the models + tools (+better harness) are better than just aider.

We tend to allow this. But you can review the diff before you allow it. It seems easier to say "roll that back" or hit esc if you see it doing things you dont like and then correcting it. I have to say the velocity is insane of coding this way. We also commit a LOT and explicitly dont allow claude to push code ever. That way we can roll back if needed, but honestly it's pretty rare to need to. The MAX plan is a must to work this way though.

1) You can plug in any model into aider 2) It can be quite agentic

> evolve tool harnesses

Claude code & Gemini cli etc. don't do this either

Seems like Amazon started making this when Cursor was hot in the market, but now that CLI agents like Claude Code are taking over, Kiro will have an uphill battle.

It’s also not free or unlimited (though throttled) like Cursor and Claude Code using max plan.

I love Claude Code too, and it definitely has it's place. I think that IDE's have a few advantages over CLI tools though. In specific the IDE has a lot more contextual information such as what files you have open, warnings from linters or type checkers, information from LSP, etc.

I think it is entirely possible to build a fantastic CLI tool for coding, and the CLI tools for coding already work well enough, but there is just more context info available inside of an IDE, therefore the ceiling is higher when working with an agent that runs inside of the IDE. Context is king for LLM results, and IDE's just have more context.

Over time I'm sure we'll see tools like Claude Code support everything that an IDE can do, but for now if you want to reach the same ceiling you still have to glue together a very custom setup with MCP tool use, and that has the downside of introducing additional tool use latency, compared to an IDE that is able to source context directly from the IDE's internal API, and provide that to the LLM nearly instantly.

Don't get me wrong, I love aider and I've been using it since the early days. I'm just commenting on the overall "gains" and imo they are higher with the latest tools (claude code, gemini, etc).

As for 1), I agree but you force the model to work within aider's constraints. Claude4 for example excels at the agentic flow and it's better at that than providing the diffs that aider expects.

As for the last sentence, I disagree. They are evolving the stack, and more importantly they are evolving both at the same time, stack + LLM. That's the main reason they all subsidise use atm, they are gathering data to improve both. If I were to place a bet right now, I'd say that provider_tool + provider_LLM > 3rd party tool + same model in the short, medium and long term.

people are trying to find a moat that will bind their userbase. Browsers, editors, apps etc. There must be a format that locks users in so they will try them all one after another

This is why you will have to pry vim and my own brain out of my cold dead hands.

It’s not just the IDE but the ML model you are selling yourself to. I see my colleagues atrophy before me. I see their tools melt in their hands. I am rapidly becoming the only person functionally capable of reason on my own. It’s very very weird.

When the model money dries up what’s going to happen?

I dunno. There's also a good chance that you just end up being left behind like graybeards that only wanted to code in C and assembler.

I too am old enough to have seen a lot of unnecessary tech change cycles, and one thing I've noticed about this industry is no matter how foolish a trend was, we almost never unwind it.

I think IDE-based tools like Cursor, VS Code, etc, will win out in the long term, especially as the younger generation grows up.

In the short term though, I think CLI-based tools like Claude Code are taking off because hardcore developers see them as the last "vestige" they have in separating themselves from the "noobs." They know there's still a good portion of the public who don't know how to use the terminal, install packages, or even know what Linux is.

I use claude code in vscode. Cmd-Esc opens a claude code tab. Then /ide conects to the vscode and it's all like cursor at that point.

I get paid a fuck load of money to write C. Your point is?

As for trends, I've been around long enough to have seen this cycle a couple of times...

> learning new key bindings

Why are they shipping them with different key bindings? Seems like the opposite of what you do to encourage product adoption.

Which is kind of ironic since the Amazon Q Developer CLI (which is essentially Claude Code with a slightly different wrapper) was released long before Claude Code and seems to mostly be flying under the radar.

gptel is great, its one of the must have packages for Emacs and I'm pretty sure that with time it will be one of the reasons to use Emacs like org-mode has been for a long time.

For people wanting to get up and running with vanilla Emacs (instead of a distribution) so that they can try out gptel sometime this week, I recommend emacs-bedrock: https://codeberg.org/ashton314/emacs-bedrock

And for a gptel backend Gemini is the fastest route (excluding something local) from generating an API key to using a LLM in Emacs (for free).

Bonus points because Emacs is useful for things other than coding you can use gptel on your notes or any buffer really to ask/talk about stuff.

> As I understand (I could be wrong), with agent mode, it edits the file for you, no way for you to accept before it does, so you have to manually check the diff, roll back parts you don't want, etc.

> Am I right?

With cursor you get reasonably flexible control at many levels. You can have it only suggest changes that you have to apply manually or you can have it make automatic changes with various ways to review, change, reject or accept. I usually have the changes made automatically but don’t accept the changes automatically. Cursor has a UI that lets you review each edit individually, for the whole file or all files. Depending on the situation I will use whichever level is appropriate. The UI also allows you to revert changes or you can ask the AI to undo or rework a change that you just approved so there’s plenty of ways to do large changes without giving up control. There’s also a stop button you can use to interrupt mid-stream if the work it’s doing isn’t what you want. It isn’t flawless but I haven’t found myself in a corner where I couldn’t get back to a happy path.

I think what is going to win is a tool independent to your ide to run your agents, it could be a cli or a gui.

The entire idea that “I’m too cool to use an IDE” I find kind of dumb. I was using a Turbo C IDE in college in 1994, Visual Studio until 2019 and since then VSCode.

Claude Code really was at the right place at the right time. Cursor started putting new models under their MAX plan that charges per use and I started getting worse results with Cursor over time as they optimized costs. I started looking into Cline/RooCode when Cursor did this because I knew they were in the squeezing customers stage now. I used those for a while with Sonnet thru OpenRouter, but Anthropic had the genius plan of bundling Claude Code with their Max plan. That made a lot of users jump ship from Cursor and the difference is night and day for me. Yes I pay 5 times more than I did with cursor, but still less than using API credits and the results for me have been superior.

They have a CLI similar to Claude Code already: https://github.com/aws/amazon-q-developer-cli

Oh, that's a good point, I misunderstood you to mean: The LLM writes it's own harnesses etc.

Same here - in fact, I just recently cancelled my Cursor subscription because Claude Code + VSCode seems just as good. I think Cursor is a decent product and some of the UX it puts around LLM interaction is helpful - but I just can't really justifying paying Cursor to middleman my requests to LLM providers when Claude Pro is $20 / month.

All of these agentic IDEs could just be visual studio code plugins. They’re likely not because how do you secure VC funding for a plugin?

You dreaming. The ui is gonna be like google for code. A voice chat and an instruction/search bar that is it. The model is the product

I don’t think “I’m too cool to use an IDE” was the point being made. The point is that having to switch IDEs every time the number one AI coding tool changes would be annoying.

Each edit as in every little change?

> I get paid a fuck load of money to write C. Your point is?

My point is there aren't many of you, are there?

All things considered, keeping up with the industry trends is generally a more reliable career path.

I just use VSCode with copilot and don't worry about these re-skins. I don't get a lot of time to write code so I certainly don't have time to learn a new IDE for a small boost to my productivity when vscode gets me most of the way already.

If these re-skinned vscode IDEs have any good ideas I'm sure Microsoft will steal them anyway.

Yes. Then you squash to make real commits.

> I dunno. There's also a good chance that you just end up being left behind like graybeards that only wanted to code in C and assembler.

All the people I know in the US with those skills make huge amounts of money- at least, the ones who haven't already retired rich.

Correct but we are never fungible. That’s the trick for a reliable career.

I’ve survived every single layoff season since 1995.

More importantly - how do you monetize a plugin?

this project was conceived when cursor was the king of vibecoding. Everyone thought being an IDE would give you so much more power.

Ironically, Claude Code has me working in lower-level languages with more low-level tools than ever before, simply because of how powerful it is, particularly as a terminal tool. I've always been more of a GUI person, but now the editor I use most often is Helix.

If I've atrophied in certain aspects of my thinking, I honestly think I've more than made up for it in learning how to engineer the context and requirements for Claude Code more effectively and to quickly dive in to fix things without taking my hands off the keyboard and leaving the terminal.

Them all being based on VS Code makes it all the more frustrating to have to switch IDEs to use them.

But Continue and a bunch of others literally did that.

The better question is why is there this horrible monoculture in SW startups around raising money through VCs? We need more regular businesses who build something useful and charge a fair price. Period.

Yes. The ones who maintain high risk system software that allows all these machines to be able to train LLMs, compile the systems software or maintaining systems that ingest 2% of the worlds traffic are the ones who are making a lot of money out of this.

Not the ones maintaining frontend web apps or "vibe coding".

I've been thinking about developing a VSIX for Visual Studio 2022 that wraps my typical ChatGPT flow in the intellisense UX.

I don't mind that everyone is all-in with VSCode now, but I already paid $500 for the big-boy version and I've got 20k hours on it.

What I like about OpenAI's codex is that it's not an IDE but a web based product that does things asynchronously from your normal workflow. It creates branches and pull requests. And you can iterate with it via git instead of working in the same tool.

So, I can use the tools I use anyway and have AIs adapt to me instead of me having to adapt to new AI powered tools. I'm using a proper IDE (intellij). Me switching to cursor, kiro, or whatever would be an enormously massive downgrade for me. These tools don't come close to the utility and features of what I am used to and depend on. And those new AI tools trying to catch up with Intellij is not their focus or roadmap. I'm not going to wait for that to happen. I need stuff that works now. Not some years after they figure it out. And that includes AI features.

There's a difference between vibe coding where you are sitting on your hands and admiring all the crazy clever stuff the AI does for you that you wouldn't be able to do yourself and working on a system that you've spent years building from scratch with AI to assist you. I do the latter. I'm constantly intervening, dismissing poor results from AI, getting frustrated with LLMs misunderstanding things, ignoring my directions, not getting the full context, etc. But I'm also getting a lot of value out of AIS with dealing with tedious/repetitive stuff, figuring out weird bugs, pointing out my mistakes, or generating perfectly usable solutions for TODOs and FIXMEs I leave in my code. About 50-60% of the PRs codex creates for me are pretty usable.

I use ChatGPT for the small stuff (it can look at intellij and apply diffs) and codex for the bigger stuff "implement foo, add some tests, and tell me when I can look at the PR". And maybe I'll check out the branch and fix a few things myself. That's something my IDE supports very well. It's not a big deal. It doesn't need to be fixed.

I find that increasingly, model quality is not the main blocker for this stuff but the developer/user experience is. Claude might be better. But chat gpt has the far better UX. And I don't even use o3 most of the time. I prefer the more rapid responses other faster models give me. It's not a cost thing but a speed thing. I only escalate to slower models when I don't like the response I'm getting. Codex is nice but slooooooow. But at least I can work on other stuff while it is doing its thing. ChatGPT gives me instant gratification. Select line, Option+shift+1, "Fix this", "....", "apply fix". That's so nice and I do that a lot. And I didn't have to replace my tools. In the same way, Claude code might be marginally better at some stuff. But the Codex developer experience is superior.

So, Kiro sounds like a nice tool for people who don't need or use IDEs. But it's not for me.

Try kilocode - https://kilocode.ai/ Its a VScode extension and allows different LLMs to be used.

Claude Code is what it is because of Claude, TUI or not isn't really the point. What makes IDEs lose to TUIs is that the agentic models can really do more than coding and is evolving toward a hands-off kind of workflow. A clunky IDE is too much for that, but TUI is not the way either. When has TUI ever been mainstream?

Agentic tools of the future will be rich notebook/chat interface that's available in all form factors, which is to say, most likely web/cross platform apps.

You can have an agent loop in your IDE, I don't see why anything makes "IDEs lose to TUIs" there.

If anything, TUIs are the awkward in-between of "human in the loop, but with poor tools" where one side is fully automatic, agents suggesting fixes on issue tracker, and the other is holding-AI's-hand where you review every step one at a time.

I hate trying to copy paste in/out of Claude Code's unnecessarily-cute boxed text input.

Zed's implementation of the agent feedback loop isn't yet as good as Claude Code, but there's nothing inherently IDE-related in the parts that are lacking.

https://zed.dev/docs/ai/agent-panel

Remaining skilled while those around you atrophy is never a bad decision

There are lots of ecosystems where pro users pay for plugins to get extra features. See the ecommerce domain, jetbrains, even ms teams has plugins that are paid.

VSCode has some popular paid plugins like LSPs or some for git.

I dont see why it wouldnt be possible to monetize a VSCode plugin.

IDEs are here to stay, I built an AI plugin for IntelliJ for that reason (https://sweep.dev/)

I use the TUI for a lot of agentic stuff that is not necessarily coding, from performing some cloud management commands or just dumb things like "where the hell did I alias vim to nvim?". For those things, having to reach out to the IDE is annoying.

And the way I see the future of coding is that should should be able to code from anywhere, mobile, web, your computer. You already have your code on the cloud (most of the time). Neither TUI or IDE works well currently for that.