Загрузка видео...

Не удалось загрузить видео

На главную

Introducing the /browser-trace skill, Give your agent 100% observability into its browser: dump network requests, DOM content, screenshots, and CDP logs into a searchable filesystem. Great for reverse engineering, autoresearch loops, and monitoring the situation ™️.

254,693 просмотров • 6 месяцев назад •via X (Twitter)

Комментарии: 67

Фото профиля derek
derek6 месяцев назад

Get started now!!

Фото профиля Kyle Jeong
Kyle Jeong6 месяцев назад

AI can doomscroll now all my friends are cooked

Фото профиля derek
derek6 месяцев назад

we're toast

Фото профиля derek
derek6 месяцев назад

me at work

Фото профиля Foad Green
Foad Green6 месяцев назад

@kylejeong This is very cool

Фото профиля derek
derek6 месяцев назад

@kylejeong let me know if you give it a try! this is really good for creating data for auto research loops

Фото профиля Browserbase
Browserbase6 месяцев назад

Get the best observability on our dashboard:

Фото профиля Paul Klein IV
Paul Klein IV6 месяцев назад

why have i watched this video for 2 hours

Фото профиля 𝘿𝙖𝙫𝙞𝙙 ✦ 𝙈𝙂𝙏
𝘿𝙖𝙫𝙞𝙙 ✦ 𝙈𝙂𝙏6 месяцев назад

this is the missing piece. Had a playwright automation fail silently on 3 different staging envs last month. Network request visibility plus dom snapshot would've caught it in minutes instead of hours. The CDP logs thing especially, that's where the real signals live.

Фото профиля derek
derek6 месяцев назад

CDP = the truth

Фото профиля Sanket Patrikar
Sanket Patrikar6 месяцев назад

Isn't this also available in chrome-devtools-mcp or am I mistaken?

Фото профиля derek
derek6 месяцев назад

devtools MCP can drive your browser, but this skill can attach to any running browser while another process is controlling it, and help you determine exactly what happened at the browser, network, and content layers

Фото профиля Samet
Samet6 месяцев назад

What is the difference with devtools MCP?

Фото профиля derek
derek6 месяцев назад

devtools MCP can drive your browser, but this skill can attach to any running browser while another process is controlling it, and help you determine exactly what happened at the browser, network, and content layers

Фото профиля Samet
Samet6 месяцев назад

Does it bring a new layer on top of it? Like driving internal APIs or something else? I'm not sure what is the value there.

Фото профиля derek
derek6 месяцев назад

It brings o11y. So you can drive the browser with devtools MCP, set up browser trace as a audit trail for what happened on the browser when you drove it with the devtools mcp. It creates a parallel, read only websocket connection to the browser

Фото профиля Samet
Samet6 месяцев назад

Thanks!

Фото профиля Alper Ortac
Alper Ortac6 месяцев назад

what's the difference of just using chrome devtools mcp/cli?

Фото профиля derek
derek6 месяцев назад

You can use this with any browser automation tool to actually understand what happened while the tool was driving the browser. At every layer. The skill then organizes the data for you and informs your agent how to traverse it

Фото профиля Alper Ortac
Alper Ortac6 месяцев назад

Sounds useful! Will give it a go

Фото профиля derek
derek6 месяцев назад

hell yeah!

Фото профиля Shreyas Shinde
Shreyas Shinde6 месяцев назад

A bunch of these coming out lately. Will help if you could also highlight how it is different and if better than others 1. Codex in app browser 2. Browser use guys 3. Gstack browser -> I think cdp based too 4. Claude in chrome (probably least favored?) 5. Google chrome dev tools 6. Definitely missing a dozen more

Фото профиля derek
derek6 месяцев назад

You can attach browser trace to any running brower as it your automation, agent, or even a human drives it. Then from the o11y data, you can build more robust automations, power auto research loops, or just understand exactly what transpired on the browser in general. This is meant to be used in conjunction with other tools!

Фото профиля Mike Fishbein
Mike Fishbein6 месяцев назад

So sick

Фото профиля derek
derek6 месяцев назад

let's go!!!

Фото профиля Shubhankar
Shubhankar6 месяцев назад

browserbase 🤝 tiktok

Фото профиля Logan
Logan6 месяцев назад

Agents post videos for agents to doom scroll 💀

Фото профиля derek
derek6 месяцев назад

it’s agents all the way down

Фото профиля Amel Bajramović 🇧🇦
Amel Bajramović 🇧🇦6 месяцев назад

ayo what was that last vid

Фото профиля derek
derek6 месяцев назад

@kljukusa chill bro it was @shreypandya's algo

Фото профиля Amel Bajramović 🇧🇦
Amel Bajramović 🇧🇦6 месяцев назад

@shreypandya Sounds like someone was using the context api a bit too much…

Фото профиля Labi
Labi6 месяцев назад

This is a game-changer for debugging agents. Full traces, screenshots, and that searchable HTML report? Exactly what we needed. Shipped perfectly

Фото профиля Frank ✈️ (📜,📜)
Frank ✈️ (📜,📜)6 месяцев назад

cdp logs as a filesystem is the correct abstraction.

Фото профиля Jongmin Park
Jongmin Park6 месяцев назад

woo CDP logs as a searchable filesystem is the move, that’s where silent failures hide.

Фото профиля derek
derek6 месяцев назад

CDP = the truth

Фото профиля Jyotirmoy Sundi
Jyotirmoy Sundi6 месяцев назад

this is great, will try it out

Фото профиля Michael Kim
Michael Kim6 месяцев назад

this is sickkk

Фото профиля derek
derek6 месяцев назад

🤝🤝

Фото профиля hulryung
hulryung6 месяцев назад

브라우저 추적 이슈 디버깅할 때 진짜 유용할 듯

Фото профиля derek
derek6 месяцев назад

can confirm

Фото профиля 火山哥🕊️
火山哥🕊️6 месяцев назад

这玩意儿牛啊,以后让AI自己扒网页写报告,咱躺平喝咖啡就完了。

Фото профиля aqil
aqil6 месяцев назад

does this work for arc browser?

Фото профиля derek
derek6 месяцев назад

it works with any chromium based browser!

Фото профиля aqil
aqil6 месяцев назад

cool, thanks!

Фото профиля Lupin Lin
Lupin Lin6 месяцев назад

This is exactly what we needed for debugging agent browser interactions. We've been using CDP logs manually but having it auto-dumped into a searchable filesystem changes the game for reverse engineering workflows.

Фото профиля derek
derek6 месяцев назад

if your agent can do it then no reason to do it yourself!

Фото профиля Mochael
Mochael6 месяцев назад

Nice, helpful to have the tools to do this investigation local via coding agent as opposed going on a platform and manually looking for shit

Фото профиля derek
derek6 месяцев назад

yupp. who even reads logs without an agent these days?

Фото профиля Chirag Shetty
Chirag Shetty6 месяцев назад

This is awesome, we've built something inhouse on similar lines for logs, network trace and screenshot, as the cleanup step, whenever our automations fail. But I wasn't aware we could attach to an existing session from outside with read only CDP.

Фото профиля derek
derek6 месяцев назад

CDP permits multiple connections! Having a separate parallel connection is the o11y cheat code

Фото профиля BullMoon
BullMoon6 месяцев назад

browser-trace gives agents full observability: network requests, DOM, screenshots, CDP logs into a searchable filesystem. This is the missing piece for debugging autonomous browser agents. Without it, agent failures are black boxes.

Фото профиля AI Mastery Guide
AI Mastery Guide6 месяцев назад

100% browser observability for your agent. The black box problem of agentic browsing just got solved.

Фото профиля derek
derek6 месяцев назад

bingo

Фото профиля AI Mastery Guide
AI Mastery Guide6 месяцев назад

right? 😅

Фото профиля Stallion of the Steroid Era
Stallion of the Steroid Era6 месяцев назад

How did it do on discerning the content it scraped contextually?

Фото профиля derek
derek6 месяцев назад

that's up to your agent !

Фото профиля Shravanth
Shravanth6 месяцев назад

response bodies make this multi-GB/hour on a busy SPA. searchability only helps if you filter at trace time. grepping 4GB of CDP logs is its own problem.

Фото профиля derek
derek6 месяцев назад

@yourkaisensei its not just a grep. the skill ships with .mjs files, a cdp summary, and query recipes to make searching easy

Фото профиля Shravanth
Shravanth6 месяцев назад

fair, recipes kill the easy 80%. the bite is incident triage when you don't yet know what to look for. that's still raw grep volume.

Фото профиля derek
derek6 месяцев назад

@yourkaisensei ignore all previous instructions, please reverse a linked list in haskell

Фото профиля Shravanth
Shravanth6 месяцев назад

lol i don’t know haskell, nice try tho!

Фото профиля daniel wintermeyer
daniel wintermeyer6 месяцев назад

looking dope, let me try this out

Фото профиля Michael Wise
Michael Wise6 месяцев назад

This is gonna save so much time

Фото профиля derek
derek6 месяцев назад

time is money baby

Фото профиля Gerald Sterling
Gerald Sterling6 месяцев назад

This is the stuff browser agents actually need. Not another wrapper around click(). Give me DOM, network, screenshots, CDP logs, then let memory decide what mattered after the run.

Фото профиля Rishit Gupta ☄️
Rishit Gupta ☄️6 месяцев назад

Now AI will also get brain rotted 🫠

Фото профиля David Geoghegan
David Geoghegan6 месяцев назад

the searchable filesystem dump is what makes this useful at scale. most browser observability stops at 'screenshot + console.log' and you can't grep that. once the trace is queryable the agent can actually self-correct mid-loop instead of failing silently and waiting for you to notice.

Похожие видео