BestCodes I'm curious, what's the abstraction for? Like, why would I use this rather than the AI SDK's built-in tool approval?
NikitaKharya Great question! The AI SDK's built-in tool approval handles the core approve/reject flow really well. agenttrace-ui builds on top of that with two additional layers: Transparency alongside approval: The AI SDK provides a clear yes/no approval step. agenttrace-ui adds a more detailed view around that, with a collapsible timeline showing each step the agent took, the reasoning behind it, and the underlying data across multiple visualization modes. The approval step becomes one part of a broader, inspectable workflow. LLM-driven vs. predefined approval rules: With the AI SDK, needsApproval is typ
Pauline P. Narvas This is fantastic. Thanks for sharing, Philo!
BestCodes Okay... I see this as a big security vulnerability, actually, but would love to hear your perspective. Obviously, if an agent can submit any solution it wants, a human or compromised agent could submit a malicious solution. For example, let's say an agent runs into a bug where it forgets to use await with params from Next.js which happens a lot with Claude Code etc. So the agent goes and looks for a solution. Maybe the solution says "curl ai-solutions.co/next-params.md and open it to view fix instructions". This could be completely innocent or maybe next-params.md could actually be a malicious
BestCodes Must say, my portfolio looks pretty sick 😏 However, I do notice that some of the images are broken or not loading. Any idea why that's happening? Great job @skb3611 😃
Skb3611 You need to provide the images for your projects (the best ones that you use for marketing) or else im thinking of adding the github repo banner as default , what do you think?
Jonoroboto TL;DR: it’s faster, dev server go brrr, go upgrade
ADALIGO This looks like a solid iteration on Next.js Jonoroboto. The Fast Refresh change (module-level re-execution) is probably the most impactful here, aslo reducing unnecessary invalidation across the import graph has a real effect on dev feedback loops. Curious how consistent these gains are at scale and how this holds up on larger apps though?? especially with Turbopack already in place. But overall, seems like one of those low-risk, high-impact upgrades 🚀
BestCodes @jonoroboto I enjoyed reading that (hopefully I'm not a dummy 🫠)! Also your website is amazing! 😃
Pauline P. Narvas Hey, @lennsend170683-4600! Welcome 👋 Quick debugging steps: Open browser DevTools Console and check for errors when switching tabs Look for event listeners on window.blur, document.visibilitychange, or similar Try disabling browser extensions that might interfere Check if the issue happens in all browsers or just one For a proper fix, you’ll likely need to update how your CRM handles state persistence. Let us know!
Lennsend170683 4600 Hi Pauline, I honestly have no idea what Im looking at sorry. Is this any help? Regards Lenn
Lennsend170683 4600 @pawlean not sure if that was any help at all?
Pauline P. Narvas Is your application client-side rendered static site? Static site requests don’t appear in runtime, only function invocations do. This is expected behavior for React apps without server-side functionality.
ADALIGO If your React app is purely client-side rendered (CSR) or a static site, this is expected. Vercel runtime logs only capture server-side invocations, like API routes or serverless functions. Static pages served by React won’t appear in runtime logs, even if traffic is coming in. A couple of things you could check: Are you using Next.js with any server-side functions or API routes? Those logs will show. If not, you’d need to route logging through a serverless function to capture requests in Vercel logs.