Editor's Note
This week brought another wave of major AI releases, with Anthropic and OpenAI expanding their agent workflow capabilities and putting pressure on traditional SaaS. Updates from Vercel and Resolve AI’s recent Series A highlight a growing focus on improving AI output in production. February is also shaping up to be a busy month for events, with amazing meetups happening across Canada and the US.
Big Trends & Ecosystem Shifts 🌎
Anthropic released 11 open-source Claude Cowork plugins that automate workflows from legal contracts to operations. This enables highly customized internal tooling and directly challenges the value of layered vertical SaaS. The release wiped roughly $280B across SaaS markets, showing how quickly AI releases can reshape entire categories.
Frontier is a platform for building and operating AI co-workers that can plan and execute multi-step tasks. Rather than just shipping models, OpenAI is moving up the stack into agent orchestration and execution. This positions Frontier as a direct counter to Anthropic’s growing agent ecosystem.
Recent case study revealed that Dropbox engineers now accept over one million lines of AI-generated code every month using Cursor. More than 90% of the engineering org uses AI weekly, and Cursor seamlessly reasons across medium size monorepos. The result is faster PR velocity and quicker understanding of complex codebases.
Developer Tools 🛠️
Vercel rebuilt v0 to close the gap between vibe-coded prototypes and production systems. The new release adds secure Snowflake and AWS integrations, a sandboxed runtime for importing full GitHub repos, and native Git workflows. It’s a clear signal that AI-first development now has to survive real enterprise constraints.
Agent Trace is an open specification for attributing AI-generated code alongside human authorship. It supports file and line-level attribution inside version-controlled codebases. As AI-written code becomes normal, this kind of provenance is becoming increasingly more valuable for teams and enterprises.
Resolve AI connects to logs, metrics, traces, and code to understand and improve AI output in production. As generating code and agents gets easier, the real challenge is making their outputs reliable once deployed. This massive fundraise aligns with current trends and enterprise initiatives to manage AI output more effectively.
Open Source Spotlight 🔍

OpenClaw (formerly Clawdbot) took the internet by storm after showing how much power a single, well-wired agent can have when it lives close to the machine. It’s a local AI assistant that runs continuously and can take real actions across your system, from browsing the web to executing shell commands and responding across Slack, iMessage, WhatsApp, and more.
The massive hype comes from how easy the tool is to adopt, and how much you can do with it. People are using it to automate research and manage all kinds of workflows locally with full control. We suggest using caution though, as giving an AI tool access to your entire laptop comes with pretty big risks.
Best Upcoming Events
🗽 New York City: Future of DevEx (February 24)
Explore how AI is changing the way we build, and get powerful insights from engineers at PostHog, Grafana, Arthur and Deskree (with a special guest from Charm).
🌉 San Francisco: Ski Camp AI: Building Agents and MCP (February 12)
An evening of demos, panels, and conversations around AI and MCP development. Join to hear from teams at Anthropic, Mintlify, Auth0, and more!
🍁 Toronto: NuSession: Engineering at Scale for 127M+ Customers (February 24)
Nubank's engineering team will share perspectives on how they design and evolve platforms at scale, and operate in distributed, high-growth environments.
Got a Question?
Got a burning engineering question? Just hit reply and we’ll tackle it. And if you enjoyed this issue, consider sharing it with your friends and teammates.
Want to Get Featured?
Want to share your dev tool, research drops, or hot takes?
Submit your story here - we review every submission and highlight the best in future issues!
Till next time,

