I'm Zach, founder of Warp, the first Agentic Development Environment (ADE). AMA! 🔥
by•
Hi Product Hunt!!
I'm Zach the CEO of @Warp , the first Agentic Development Environment (ADE)..
We are in the midst of a change from "development by hand" to "development by prompt," where instead of hand-writing code and commands, developers ask agents to build features, fix bugs, debug server crashes, and more. At Warp, we are building the first agentic development environment designed from the ground up for this new workflow of humans and agents collaborating to ship better software.
Ask me anything about how we are building Warp (e.g. why it's a native rust app), why a new environment that's neither a terminal nor an IDE is needed, how cloud agents fit into the picture, and where this is all headed in the next few years."
— Zach
1.7K views
Replies
Coming from Cursor, I've been surprised to see no counter of current context length of each conversation window. What is Warp's idea on tracking context length and resetting / summarizing context?
Warp
@gusto_js We actually recently added something that shows you when the context window starts to fill up. It's a little battery icon that you can hover over and see the context window usage. It shows up in the prompt. I'm trying to find an image of it.
We also added a second feature recently that uses an LLM to suggest when you should start a new conversation. So it kicks in when you change subject midway through.
Finally, our general approach is that we try to make users not have to think about this too much and we will do things like summarize and truncate for you if you don't actively manage yourself.
Hi @zach_lloyd
I have been using Warp for 2 months now, and I really appreaciate how it integrates agent/llm asking with command line.
I confess I don't use it specifically to code, since I delegate it to Claude Code and Cursor.
But I use it a lot to quickly understand a part of codebase, configuration files, and the most used: to generate complicated commands using natural language.
The models that are offered by Warp today are the "mainstream" commercial models (Sonnet/Opus/Gpt, etc) but they are kind of slow, which take me out of the flow when executing some cli actions/devops/etc
I've recently published in twitter a simple alias to use a fast llm inference like Groq to generate commands from natural language
I always give the aws cli examples because it requires a bunch of parameters depending which service I'm using. Asking things like: "give me all the records on my meetings table on dynamo that lenght is greather than x, give me it in json format".
This ask, in the normal Warp agent using the currently offered models, would take at least 30 seconds and also it will come with explanations, etc.
I just was to run that command, but asking in natural language as above.
So, integrating fast inference/llm, would make users achieve this without getting out of the flow.
With that said, I would like to know if you guys have any plans to implement fast inference models/servers that allow us to use Warp as a "natural language" command line for every commands.
Warp
@douglas_correa Really good question. Couple thoughts here:
1. You can pick a faster, cheaper model to do things in Warp. So I would start with that by changing the model in the model selector to something that's lower latency, maybe like GPT-4.1. We are also thinking about implementing some model routing so this can happen automatically.
2. For the real-time translation of English into commands, we actually have a feature that's built for this that you can activate by typing # into the input and we'll bring up real-time English to command translation. It might not have exactly the latency that you're looking for, but you should try it and give me feedback.
Dashcam
How do y'all test the Warp client?
Warp
@sw1tch We have a whole bunch of different ways of testing it: everything from unit tests to an integration test harness that actually runs the full app and simulates user actions to dogfooding it to putting it out on a preview build to being able to run different experiments where certain features are enabled for different parts of the user base. So we test it at a lot of different levels.
Hey Zach, huge fan of what you're building with the ADE.
Thinking ahead to more complex workflows, where you have multiple agents collaborating (e.g. one scaffolds a service, another containerizes it, a third deploys)... My main question is about the handoff: How do you see agents sharing state and managing dependencies between each other? Is that something you envision Warp handling natively, or more through ecosystem tools?
Warp
@dhairya_thakkar1 This is an awesome question. I think we'll probably support both, so we'll want to make native handoff between contexts between agents seamless in Warp.
But we also support things like MCP for this sort of thing. We're also launching a Warp MD file that's compatible with Claude MD and cursor rules. So it's kind of about both approaches being in play.
@zach_lloyd Makes a ton of sense to have both Warp-native handoff and MCP/Warp MD in play. Passing context cleanly feels like (at least to me) what might fully unlock multi-agent workflows for bigger tasks.
On a side note, curious what you and the Warp team look for in early-career folks (new grads, interns, etc.) whether that’s particular skills, attitudes, or just the kinds of problems you like to see them dive into.
Invoice generator
Hey Zach! Any plans for Grok 4 integration? Users claim that on X, it generates much less bloated code.
Warp
@csaba_kissi We have explored Grok 4 but haven't yet done the work to ship it. Every time we ship a new model in Warp, we want to make sure that it's high quality, which has to do not just with the underlying model but also how we prompt it, how we do the tool calling and prompt caching.
There's a whole bunch of work that has to go in, and then we actually want to run our internal evals on it to make sure that it's good for users. So, short answer is, we are evaluating every new model that's coming out.
We haven't quite gotten Grok 4 over the line, it's possible that we'll launch it, and we definitely want to continue to offer people who are using Warp state-of-the-art models from all the providers.
Hey @zach_lloyd , long time user of Warp here, and I had the chance to use Warp 2.0 as soon as it was released.
Before that, for my agentic developments, I used a lot of Windsurf/Cursor/Copilot. I was switching between them because the cost of switching was very low so I was using the best IDE at any point in time.
I recently switched my development workflow totally to Claude Code (well, I still open the project in VSCode for the diff and the review haha).
Today I'm still considering the switch from Claude Code to Warp, but I still have a few questions:
- what are your differenciators vs Claude Code, aka what can you do with Warp that you cannot with Claude Code?
- I have the feeling that I can still ask Claude to execute terminal commands and ask it to debug the results, correct? If yes, then what are the advantages of Warp?
- There are no agents/subagents yet in Warp, do you plan to develop these kind of features? Same question for the /commands.
- How is the performance of your agent compared to Claude Code that uses the Claude models "fully"? I'm using Claude Code with AWS Bedrock (not a Claude subscription) so I pay for every token I use, so the development performance is very powerful. How do you compare to that as you're on a subscription model?
thank you
Warp
@benjamin_sicard This is a great question, and I'm glad you're a fan of Warp! The short answer is, it's totally cool to use Cloud Code within Warp, but we feel like we have several advantages.
The biggest one is that Warp, by virtue of being a GUI app, can do things at the UI layer that you simply can't do with a purely TUI-based app like Claude Code. If you check out our preview build, you'll see it's possible to review code diffs, edit the diffs produced by the agent, and iterate in a much tighter way than if you do your work with Claude Code, where it ends up being that you have to context-switch into either an editor or the GitHub code review UI or something like Git Tower to actually see and edit what the agent is doing.
I think this is really important if you're doing production coding. It's less important if you're doing vibe coding, where you don't have strong requirements about the maintainability of the code or maybe you don't need to comprehend it at the same level.
We also at Warp have the advantage of being able to be multi-model, so that the next time a model comes out that we think could either be either better than Claude or could supplement Claude and be used in a different way. You're not stuck with one model provider, and you don't have to redo all your workflows to switch across model providers.
For subagents, I'd be really interested in hearing what your use case is. It's something that we're considering adding in Warp But I'm not yet fully convinced of the value for most users, but I could potentially get there.
For the pricing and performance, we are, I think, at a par basically with Claude Code's efficiency. We're not trying to compete on price. In fact, I think that's a losing strategy for us. We're trying to add value through the application layer on top of what these models provide.
Nice, Really like this. Lowkey wild to think we might look back at hand-coding the way people look at writing assembly today 😂.
Warp
@pravin_meh66451 I like that analogy. I do think we're moving up one layer of abstraction. It used to be assembler to programming languages, and now it's kind of programming languages to English.
Transitioning from Cursor/Code to Warp - with large code-bases. Would really like the ability to have a file-browser on the side panel - remembering all the folder structure and filenames is hard without it, and something coming from a IDE thats hard to release, when you want to target the prompt to consider specific files for amending.
I still find that Warp 'forgets' context between long sessons - and i have to remind it to look for .md files for consideration (memory.md or warp.md etc) is there any way to force items into the index?
Is the index local to the machine or in the [warp] cloud - if i use warp between laptop and desktop will it remember its state.
Keep up the great work guys!
First up. I subscribe to and LOVE Warp. Its the single AI tool that has made the biggest difference in my work- no more trying to remember an obscure command line tool name that's riiiiiight there.
My question is, how far are you planning to build Warp out? Right now it is my go-to terminal client, since its UI is fantastic, and it helps me get work done faster.
Are there plans for key vaults/ssh shortcuts like Termius, or other features targeted towards sysadmins?
Perhaps a more fleshed out memory/guardrail/agent management system, allowing me to easily manage my claude code instance from Warp?
No wrong answers as far as I'm concerned, just interested to hear your plans, and excited at the thought of replacing other tools and workflows with Warp