AI Agents
The Agentic Software Factory: Every Control Surface, Named
Thirteen pieces built the parts. This one puts them together: the whole flow from intent to verified behavior, the nine control surfaces where humans …
You're Not Managing Engineers Anymore. You're Allocating Judgment.
Sprint planning handed out the scarce resource: engineer-hours. That resource now stretches on demand, and the thing that is really capped is how much …
Your Agent Has Prod Access. Your IAM Thinks It's a Person.
Every identity system we built authorizes a principal making a request, assuming that principal decided to make it. An agent’s request comes …
DORA Measured the Old Constraint. Agents Broke the Instrument.
Deployment frequency and lead time worked for one reason: they were hard to fake while humans typed every line. Agents made them easy to inflate. …
Technical Debt Just Got Cheaper. Something Worse Moved In.
The migration you put off for six years is now a weekend. Classic technical debt got far cheaper to pay off, because mechanical change stopped costing …
Your Dashboards Were Built for Eyes. Your Operator Doesn't Have Any.
·14 mins
A dashboard throws most of your telemetry away, and it is tuned for what a pair of eyes can take in. Alert thresholds exist because people cannot …
Stop Asking Whether Agents Should Act in an Incident. Decide How Far.
Break down your MTTR honestly and most of it is human delay: paging, waking, getting your bearings, forming a theory. Agents remove exactly that …
Your Infrastructure Assumes a Human at the Keyboard
·15 mins
Terraform, Kubernetes, and every cloud console were designed around a human who hesitates, notices the environment name, and feels dread before typing …
Your Pipeline Runs Scripts. It Should Be Making Decisions.
CI/CD was built to run a fixed sequence for a change a human already understood. Under agent volume that design fails in one specific way: the same …
Coverage Was Always Theater. Agents Made It Free.
When an agent writes both the code and the tests, you get a closed loop with nothing checking it from outside — and a coverage number that goes up. …
Human Review Is Scarce Now. Stop Spending It Line by Line.
Do the maths on your review capacity and the answer is brutal: most teams passed it long ago and are covering the gap with approval theater. …
The Scarce Skill Isn't Writing Code. It's Knowing What Shouldn't Exist.
Seniors who defend their value by pointing at code they wrote are arguing from ground that has already gone. The engineers who win the next five years …
A Ticket Is a Lossy Compression of Intent
Tickets worked because a human unpacked them using context nobody ever wrote down. Hand the same ticket to an agent and you find out how much of your …
Rethinking the SDLC: Execution Is No Longer the Constraint
The software development lifecycle is a stock-control system for a resource that stopped being scarce. Every phase gate, every batch, every handoff …
Make the Cheap Path the Default. Make the Expensive Path Prove It.
Half of most AI agent bills is waste with a good story: context pollution, verification theater, tool schemas nobody reads, retries nobody capped. …
Alberta Scanned 466 Million Lines of Code in 20 Hours. The Architecture Is the Story.
A Canadian province ran Claude Code across 466 million lines of government code in 20 hours — work they priced at 6.5 years the old way. The number …
Stop Reviewing Code. Start Reviewing Evidence.
Agent-heavy teams ship twice the PRs at more than twice the size, and reviews wait almost five times longer for pickup. The review model built for …
Agentic Overwatch: Why Your Next Dev Team Will Look Like a NASA Control Room
·3104 words·15 mins
Agents don’t just write code anymore. They run ops, security, QA, data, and support, around the clock, while we still govern them with a team …
The One-Man Show Company. Don't Let the Monkeys Touch Production.
·2834 words·14 mins
A company used to start with people. Now it can start with one person and a swarm of AI agents that draft, build, test, sell, and support faster than …
The End of Courses: Learn From AI Like a Toddler, Or Become Obsolete
·2517 words·12 mins
Remember when shipping an app meant 40 hours of video courses and weeks of syntax memorization? An agent builds it in three minutes now. The 40-hour …
I Don't Put All My Eggs in One Basket. Anthropic Is Making That Hard.
·2181 words·11 mins
Anthropic shipped 120+ features in 90 days, then blocked OpenClaw from using Claude subscriptions. The same company building the best developer tools …
Claude Can Now Use Your Computer. Here's What That Actually Means.
·826 words·4 mins
Anthropic just shipped computer use for Claude. It can click, scroll, navigate your browser, open files, run dev tools, and submit PRs. Pair it with …
DeerFlow 2.0: ByteDance Just Open-Sourced What Most Companies Are Trying to Build Internally
·1149 words·6 mins
37,000 GitHub stars in weeks. #1 on GitHub Trending. ByteDance rebuilt DeerFlow from scratch into a super agent harness with sandboxed execution, …
Zuckerberg Is Building an AI CEO Assistant. The Rest of Us Should Have Started Already.
·573 words·3 mins
Mark Zuckerberg is reportedly building an AI agent to help with his CEO duties. My reaction: this is obvious, and frankly late. I’ve been …
OpenClaw Is Not a Chatbot. It's a Personal Agent Gateway.
·1907 words·9 mins
Everyone keeps comparing OpenClaw to ChatGPT. They’re looking at the wrong layer. OpenClaw isn’t trying to be a better chat UI. It’s …
Your AI Agents Are Flying Blind. Here's How to Fix That.
·2733 words·13 mins
Every AI agent in your org starts every session with zero context. No business rules. No architecture decisions. No conventions. The code they …
Org Charts for AI Agents: Mapping Your Human and AI Workforce
·3193 words·15 mins
AI agents aren’t coming. They’re already here, doing real work, while most organizations are still debating how to use ChatGPT. If …
