The least productive thing I can imagine doing is watching an agent work. But I still do it, because I'm never really sure if the agent understood what I wanted it to do. So I end up reading the tool calls and watching files change. I want to agree on what it's gonna build, leave it alone, and review what comes back. But an all-text plan that's several pages long just feels like homework, and it's easier to let the agent run, even if it wastes tokens, just so I can see the result. What I need is a better way to verify the plan and review the result, just at a glance. Something where I can see the layout or the API flow and point to the part that's wrong. Here's what I've been doing instead lately. builder.io/blog/stop-watchin…
5
15
2,039
I keep getting frustrated when I ask an agent to clean up some code and get another wrapper, another helper, and all the original complexity still sitting underneath. Then I have to figure out if anything actually got better. We’re running into this at Builder while developing Agent Native pretty much as fast as we can. People leave feedback in Slack, and an agent factory picks it up and works on it. A lot is changing, all the time. So validation is a big deal for us. In one cleanup experiment, an agent removed a lint warning, passed typechecking, and passed all 23 tests in the focused suite. The code still used unvalidated input. Everything passed, but we still hadn’t checked the behavior we actually needed. I wrote up what we’re learning about asking agents to clean up code, figuring out what needs to change, and checking whether it actually got better. builder.io/blog/de-slop-ai-g…
3
6
1,133
Live now. Building a plan from scratch, reviewing a real PR, triaging a ticket — all in Plans. free and open source → luma.com/plans
4
865
Sept 15: we build a plan from a blank prompt, review a real PR the same way, and triage a ticket like an agent would. 20 min, and a free, open-source tool you can use today → luma.com/plans
3
711
Builder.io retweeted
Introducing /webmcp - make your agent workflows visual
6
13
174
15,429
Plans turns your coding agent's plan into wireframes, diagrams, and annotated code you can actually review. Comment on any block, the agent revises. free and open source. live demo sept 15 → luma.com/plans
1
2
655
You shouldn't have to talk to 100 people to build one deck. Connect your sources — Slides builds it on-brand, and because it's an agent, you can prompt it to rebuild itself on any interval you'd like.
1
1
1
604
Reading your coding agent's plans is rarely part of the plan. It's a wall of markdown in your terminal, gone the second you approve it. There's a smarter way to work. free, open source. live demo sept 15 ->luma.com/plans
1
577
Builder.io retweeted
Loom records the crime scene. It doesn’t collect the fingerprints. Agent-Native Clips records the bug AND captures console errors, failed requests, and status codes from the same session. Free. Open source. Built for humans and coding agents.
2
6
14
13,712
Gamma creates a decent deck fast. Then you're in its canvas, on its subscription, hitting regenerate when it's 85% right. Agent-Native Slides is free, open source, knows your brand, and lets you edit anything. See it live Sept 1. luma.com/slides
2
2
5
1,760
Ask Claude or ChatGPT for a deck and you get a file. Numbers change, the file's useless. So we built slides that build themselves: when the numbers change, the deck updates instead of making you start over. Agent-Native Slides, free and open source. Sept 1, live. luma.com/slides
2
734
You make a deck with AI. It hands you a file and walks off. Numbers change, you're back to square one. Sept 1, live: watch a CSV and a Slack channel turn into a real exec deck, the kind someone on your team rebuilds by hand every week. That's Agent-Native Slides, free and open source. luma.com/slides
3
6
673
Most decks die as an email attachment you can't track. This one ships as a live link. It updates when the numbers do, and you can see who actually opened it. Agent-Native Slides. Sept 1, live, free and open source. luma.com/slides
1
3
1,562
"AI decks look like AI made them" is a fair complaint. So we didn't leave brand to the model. Point it at a deck your designer already made and it builds in that style: layouts, headings, colors. That's Agent-Native Slides, free and open source. See it live Sept 1. luma.com/slides
1
663