Video yükleniyor...
Video Yüklenemedi
1/ Introducing Cline Cloud Agents. You can now use Cline straight from your browser. Give it a task and it runs in a secure cloud sandbox, writing code, testing its work, then opening a PR on your GitHub repo.
12,709 görüntüleme • 3 gün önce •via X (Twitter)
11 Yorum

2/ Run several Clines at once in parallel. Every agent gets its own independent, secure container in the cloud. You can hand off a bug fix, a refactor and a new feature at the same time, then review the PRs as they come in.

3/ It works on your phone too. Start a task on your laptop, close the lid, and check in from your phone. The agent keeps working while you're away.

4/ Use any model. You can try it for free with models like DeepSeek-V4.1-Flash. Or use ClinePass, our subscription for open-weights models at ~5x discounted access.

5/ To get started: 1. Log in or create a Cline account 2. Connect your GitHub 3. Open the Agents tab, pick a repo, and tell Cline what to build

You guys are killing it.

The interesting move here isn’t “agents” — it’s that the environment is the browser itself. Most coding agents live in the terminal. How sticky is the state though? If I queue a task tonight and check tomorrow morning, does it actually remember where it left off?

Hey bro!Could you please consider adding Claude models for Cline Pass users? Even just Haiku 5.5 would be great, especially since competing tools like OpenCode and Command Go already support it. Without it, I'm seriously thinking about dropping my Cline Pass subscription.

How does the cloud sandbox handle custom environment setups or private dependencies before running tests?

very nice thanks guys for this banger

it is showing this nav, not multiple sessions/agents parallel as in this video ( or like in Cline Desktop ) , am i missing something, or pls fix this

@cline This is the right default for cloud agents: code, test, and PR from a sandbox instead of a chat transcript. The missing receipt I look for is boring: repo, diff, tests run, failure path, and who can replay the run without trusting the agent log.
