Tech Skills
A coding agent turned loose on a tangled monolith behaves exactly like a junior who’s never seen the codebase: it guesses, it duplicates, it confidently wires the wrong things toge
When you add an MCP server, install a skill, or drop in a plugin, it feels like editing a config file. It isn’t. You are installing code and instructions that run with your agent’s
Here’s the most uncomfortable software finding of 2026: reviewers feel better approving AI-generated pull requests than human ones — even though agent code carries roughly 1.7× the
Your CI dashboard says 92% coverage on the feature your agent just wrote. It feels like a safety net. It isn’t. Line coverage measures whether your tests executed a piece of code —
Point a single AI agent at a million-line codebase and watch it slowly lose its mind. It forgets the decision it made twenty steps ago, poisons its own context window with the outp
For decades, the fastest typist with the best memory for syntax looked like the best engineer. That era is over. When a fleet of AI agents can write more correct code in an afterno
When agents write most of your code, the pull request stops being a defect-catching gate and becomes something more important: the steering wheel of the entire system. The compiler
An agentic loop that no one measures is a slot machine. It feels productive — tokens are flowing, PRs are opening, dashboards are green — but you have no idea whether the machine i
When you go from one engineer running one AI agent to a fleet of code-writing agents running an enterprise loop, you haven’t just added a productivity tool. You’ve stood up a new p
An agentic loop is a slop amplifier by default. The same fleet of agents that lets your team ship at 10x will, left ungoverned, turn a ten-million-line codebase into an unreviewabl
