Loading video...
Video Failed to Load
Hello!, we've been silent for a while. A thread🧵: What we have shipped and what we are shipping today Today, we announce that we're upgrading CranL to be fully Agentic. Where your Agent writes code, commits and PRs. Create an account in CranL And deploy. We're Agent-Ready since our... show more
40,260 views • 4 days ago •via X (Twitter)
18 Comments

Also. Today we're releasing: CranL Realtime📶 A feature that allows you to: - sync client state over WebSocket - Listen to database changes - Store users status - broadcast messages All that and more comes fully managed from CranL. With fully API and MCP/CLI access. To help you scale in few milliseconds!

As we continue, Today we are publishing: CranL Cloud Agents 🤖 A place where you can run your coding Agent inside CranL. Or just tell Claude Code/codex/Grok to work there on your prompts. Cloud agent will perform your prompt. Write code. Submit a PR. Let you know and die. Billed as you go, and easy to use!

As well as, CranL Sandboxes are officially out of Beta Now you have full control over your Sandbox, billed as you go. Short-living and long-living. With fully API/MCP/CLI Access. Which allows you to work seamlessly and easily.

And last but not least, we are rolling out the new UI/UX of CranL. Where we invested a lot of time into making the experience better, cleaner, smoother and easier. Along with thoughtful and clean designs. A huge upgrade to our platform!

And that's it, you can now try CranL for free too (:

قوي

وربي قوي

This is just amazing 🔥

احبككم

اعششششق

Amazing work 🔥

🔥🔥🔥🔥

Best tech company ever ☝🏿🥇

An MCP/CLI with 450+ actions is like giving an agent a full set of keys; the important part is deciding which doors it can open without approval.

ماشاء الله تبارك الرحمن، بالتوفيق باذن الله

Congraaaaats 💕💕🎉🚀🚀

Is 450 the count of exposed endpoints, or what one agent run can actually finish?

Cranlcom needs a tool-call receipt: server, tool name, schema version, auth scope, input hash, output hash, refusal reason, and replay state. Agents can reuse the result without trusting the chat transcript.

