GPT-5.2 Lands This Week & Linux Foundation Opens Door to AI

Hello!

OpenAI’s code red is pushing GPT-5.2 into this week’s spotlight as the race with Gemini 3 picks up speed. Over at Meta, things are a bit more unsettled, with strategy shifts and internal friction shaping the conversation. Anthropic, on the other hand, is bringing Claude Code into Slack.

If you’re working with React or Next.js, the React2Shell vulnerability is worth a close look. And in the wider ecosystem, the Linux Foundation just launched a new Agentic AI initiative that immediately pulled in almost every major player.

Sounds good? Grab your coffee and enjoy Frictionless!

In the Queue

• A Smarter Path to Legacy Migration

• Why Estimates Keep Breaking Teams

• The Google Engineer Who Avoids the Spotlight

• Mapping the Relationships Your Org Runs On

• DoorDash’s 80M RPS Service Mesh Playbook

• React Hooks Need an Upgrade

• How to Build an AI-Native Engineering Team

Reduce Friction

Why Labeling Relationships Is So Important

Many projects fail because no one maps the relationships between the pieces. Check how goals, initiatives, bets, and feedback loops connect. You'll see your company's real operating system is far more complex than any org chart suggests.

Estimates – A Necessary Evil?

Product owners need estimates to manage deadlines and release plans. But developers are being asked to predict work in systems riddled with technical debt and hidden dependencies. The real problem isn't the estimate, but how fast it becomes a commitment nobody wants.

Why One Google Staff Engineer Avoids the Spotlight

Deep, long-term ownership often creates more impact than chasing high-visibility projects. This is especially true in infrastructure, where familiarity and experience are a must. Some of the most meaningful careers are built by staying with a system long enough to shape it.

Unlocking High Engineering Pace: Prevent Unplanned Work

When 30–40% of your team's time goes to outages, defects, and tech debt, you're effectively losing a quarter of your roadmap. Reclaim your velocity by tracking interruptions and eliminating the failures you can directly control.

Deepen Your Expertise

React2Shell: Technical Deep-Dive & In-the-Wild Exploitation of CVE-2025-55182

If you're running React Server Components in production, React2Shell needs your immediate attention. Attackers are exploiting CVE-2025-55182 to execute remote code on Next.js and other RSC apps, deploy crypto miners, and drop Sliver backdoors into Kubernetes workloads.

Strategy for a Successful Legacy System Migration to Next.js

Legacy migrations might seem overwhelming, but Joanna shows that it doesn’t have to be the case. She surfaces the hidden costs teams often miss, and then walks you through a clear path to safe modernizing with Next.js.

React Has Changed, Your Hooks Should Too

React has moved beyond treating useEffect as the default answer for everything. Newer patterns like useSyncExternalStore for subscriptions and transitions for concurrency help you build components that behave more predictably.

How DoorDash Moved to a Service Mesh to Handle 80M Requests/Second

DoorDash standardized reliability patterns across 1,000+ services by building a custom Envoy mesh that handles retries, load shedding, and traffic control. The strategy was to ship a bare-bones MVP quickly, then automate everything to support 80M requests per second without making reliability every team's problem.

AI Corner

Meta’s Shifting AI Strategy is Causing Internal Confusion

Meta’s AI efforts are starting to show the strain. The open-source strategy is losing momentum, and new leaders are pushing a culture that the rest of the company isn’t happy about. To make things worse, while Meta is trying to find its footing, their competitors are racing ahead.

OpenAI Guide: Building an AI-Native Engineering Team

OpenAI's guide on building an AI-native engineering org explores delegating entire workflows to agents that can execute tools across your stack. Teams using this approach are engineering differently: humans keep the hard decisions while agents handle the first pass on everything else.

Claude Code is Coming to Slack, and That’s a Bigger Deal Than It Sounds

Claude Code in Slack shows AI moving from improving developer workflows to becoming the workflow itself. When coding sessions, PRs, and bug fixes happen inside chat threads instead of IDEs, controlling where developers work matters more than having the best AI model.

OpenAI Races Gemini 3 to the Top With GPT-5.2 Drop This Week

OpenAI is fast-tracking GPT-5.2 after Gemini 3 drew strong reviews. The focus is on making ChatGPT faster, more stable, and better at reasoning. Originally planned for later this month, internal pressure moved the release forward, with OpenAI shelving experimental projects to improve performance.

Just Cool

Linux Foundation announces Agentic AI Foundation joined by Anthropic, OpenAI, Block

The Linux Foundation launched the Agentic AI Foundation, backed by Anthropic, OpenAI, Block, and nearly every major cloud vendor. With MCP, AGENTS.md, and Goose now under neutral governance, the industry is working to standardize a space where adoption is growing fast, but enterprises are dealing with operational risk and inconsistent returns.

Let’s Stay in Touch! 📨 

Do you have any comments about this newsletter issue or questions you want to ask? Drop me a message or book a meeting.