Switch into Agent Mode and accomplish any dev task from your terminal. Agent Mode understands natural language, can execute commands with your permission, is self-correcting, and can learn any CLI. Try Agent Mode in Warp today.
Hey Product Hunt 👋
David, product designer at Warp here — Today, we’re super pumped to announce the launch of Agent Mode in Warp AI.
When we set out to build Agent Mode, we wanted to imagine what an AI-native, deeply integrated terminal experience might look like.
What we came up with is a new way to interact with the command line using natural language. What makes it extra powerful is that if Agent Mode needs more context to complete your task, it can suggest commands to gather more context, read the output, and use that information to help you achieve your goal.
We believe this short feedback loop between developers, AI, and commands is a powerful interaction paradigm for software development.
Privacy and transparency were key principles guiding the design of Agent Mode. If Agent Mode needs to run a command, it will always ask your permission before doing so. In addition to giving the user visibility into what information is being sent to the LLM, it also creates transparency in the system, allowing you to intuit how it works.
We’re really excited to share Agent Mode with everyone here at Product Hunt, and developers everywhere.
To give Agent Mode a spin, download or update Warp today and just type “hello.”
Interesting idea. I think the terminal along with ai is very powerful. Congrats on the launch!
Hey David and the Warp team! 👋
Congratulations on launching Agent Mode in Warp AI! As a Tech Recruiter Lead, I'm intrigued by innovations that streamline developer workflows. Agent Mode's integration of natural language with the command line promises a more intuitive and efficient coding experience.
I'm curious, how does Agent Mode handle complex tasks that require multiple commands and iterations? Excited to see this in action!
I had a chance to preview this functionality about a month ago, and was really blown away. Something I could never imagine possible, having plane-text help to help me work through, often challenging, command line tasks - this is awesome! I find that command line interface is a natural UI for LLM functionality and the Warp team did a great job integrating it seamlessly into the experience, love it! 🌪️
Are there any collaboration features within Agent Mode for teams working on the same project? How can multiple developers benefit from this AI integration?
Great launch!
Agent Mode is a truly interesting way to interact with the command line. Congrats on the launch, Warp team!
Love the concept, David! How do you see Agent Mode evolving in the future? Are there any additional features or integrations planned for upcoming updates?
Hey guys, congrats on the launch. Seems like a really cool project with Devin and other products inspired by it growing exponentially. How would you say that this is different from the former? And is the entire LLM in our local machine or server is involved even after setting up?
I'd just be happy if warp worked at all using a fish shell in Linux. Issue has been open for months and months and every new release I hope it might work... but it doesn't.
Hi all. Just wanted to give some feedback on warp/AI. I downloaded the tool, and started trying it, and while its great, I found an issue. I was asking questions about extracting mp3 metadata, and it suggested to use the tool exif. This is fine, however, when I typed in exif -h to get more information, I accidentally typed exit instead, and I lost the window and it hadn't preserved my history. it might be a useful option to handle that sort of thing, in settings perhaps? I do realize there's several ways I can handle this, including putting a wrapper around the tool, or maybe renaming the binary.
