Happy Friday Eve, legends! Welcome back to the Leaderboard. In today's edition, we're taking a look at an AI that could automate your work, a new image model that doesn't output slop, and a C interpreter that could make college a lot easier. Let's dive in.
Automate your work
Foundry: Build, test, and improve AI agents
There’s endless talk about how AI will radically change the nature of work, but what will that actually look like in practice? I think Foundry gives us an idea. The platform lets you build and tweak custom agents that can do everything from customer support to internal ops. But these agents aren’t wind-up dolls; they require frequent inputs and tweaks from thoughtful humans. In other words, humans in the workplace aren’t going anywhere — they’ll probably just spend a lot of their time supervising armies of agents.
Pulling teeth
Runway Frames:Â Runway's latest foundation model for image generation.
Getting Gen AI to create consistent, tasteful images usually feels like pulling teeth—like trying to guide a grade schooler through an art project, hoping for something that’s at least passable. Frames from Runway changes the game. It’s the first time it feels like the AI actually gets it. You’re not babysitting or endlessly tweaking; you’re stepping into the role of an Art Director. You focus on the big picture—world-building, creative vision—and the AI follows your lead, delivering results that actually feel aligned. It’s less of a chore, more of a collaboration, and honestly, it’s a huge relief.
Your agents stopped answering and started doing
Your agents call tools, write to databases and move money, so a wrong answer is now a wrong action, which is a considerably worse meeting. Arcjet runs policy before every LLM call, tool call, query and API request: prompt injection, sensitive data, bots, rate limits, spend quotas, and your own rules in Rego on top. It reads the OpenTelemetry you already emit, so switching it on is one environment variable, no code changes and nothing to deploy. Each check also reads the earlier steps in the same workflow, so an agent's tenth action gets judged against the first nine and not on its own. JavaScript, Python and Go SDKs all at 1.0, agent framework guards included, free tier to start on.
Students rejoice
CJIT: A C interpreter that lets you run C instantly.
Shortening the feedback loop has obvious implications for the velocity of any development team, and CJIT promises to do just that for folks working in C that spend their days waiting for their code to compile. I'm slightly scarred by my experience with lower-level languages from time at university and really would've appreciated this kind of tooling then -- I hope this helps all the warriors using C today.








