@junhssss
Joined June 2024
  • Tweets19
  • Following4
  • Followers9
Most browser-agent problems show up after the happy path. Run two. Same site. Same account. New IP. Lost session. Unclear spend. Trace that does not tell you enough. Everything we shipped last week to solve this. ↓
3
2
1
14
691
Launch week -> Day 05 Today, we're rolling out simpler pricing. Launch, Scale, Enterprise. Metered browser rates, usage credits, visible limits, and a usage dashboard that shows spend before the bill arrives. 🧵
6
9
4
29
2,735
Launch week → Day 04 Today we're open-sourcing Atlas. A deep research harness you can run, fork, and bend around your own domain. 🧵 h/t @junhssss
10
8
4
40
3,559
Launch week → Day 03 Today we are shipping native Rust and Go SDKs. Spin up cloud browsers natively in the stacks your agent infrastructure runs in. 🧵 h/t @junhssss
9
6
4
32
2,957
Launch week → Day 02 Today, we're rolling out Dedicated IPs A stable network origin for agents that log in once and need to keep running from the same place. 🧵
13
7
6
29
3,661
Launch week → Day 01 Today, we're rolling out Stealth Browser Our own fork of Chromium - fortified at the source code level to access the hardest sites to bypass on the web. 🧵
10
11
7
58
9,382
5 days. 5 drops. We're not saying what they are yet. Steel Launch Week starts Monday. @hussufo explains 👇
9
8
2
37
4,342
this was fun building with Nas
nars
@0xbostaJun 5

Skills aren't prompts

Article · @0xbosta
1
1
134
Jun retweeted
nars
@0xbostaJun 5

Skills aren't prompts

Article · @0xbosta
5
3
5
12
5,022
Stop paying for a browser after the work is done or your agent dies. Steel sessions now take an inactivity timeout: if the agent stops driving the browser for the window you set, Steel releases it and the meter stops. Read more ↓
2
2
6
269
Most of what you ship to an agent gets compressed before it acts — docs, SDKs, blog posts, distilled by the model first. Errors are the exception. They reach the agent intact. We've been rebuilding ours around that. ↓
1
2
2
7
457
the practitioner angle If you're building browser or computer-use agents, this is the fastest way to see who's actually ahead, and how each number was scored. Read past the top row.
Steel
@steeldotdevMay 29
Browser-agent benchmarks are getting crowded, stale, and hard to compare across. We are collecting the benchmarks that actually matter for browser and computer-use agents, so you don't have to chase them down. (Just rebuilt the whole leaderboard. Live now.)
1
4
234
Browser-agent benchmarks are getting crowded, stale, and hard to compare across. We are collecting the benchmarks that actually matter for browser and computer-use agents, so you don't have to chase them down. (Just rebuilt the whole leaderboard. Live now.)
5
3
5
15
3,045
Load more