Hey there! We've got a fresh update ready for you with some meaningful improvements to how Juma handles AI-powered tools and agent workflows. Here's what's new:
We've bumped the Shell output limit to 500 lines, so you can now see much more of what's happening when agents run commands - with clean UI truncation so things never feel overwhelming.
On top of that, large MCP integration responses are now handled more efficiently behind the scenes, meaning complex agent tasks are more reliable and less likely to hit snags.
As always, if anything feels off or you have feedback, we're all ears. Happy building! 🚀