Posts

I Added an AI to Every PR for 30 Days. Here's What Actually Caught Bugs

I added CodeRabbit to our GitHub org on a Tuesday. By Friday, it had flagged 23 issues across 7 PRs — and my team thought I had secretly reviewed everything myself. After 30 days running four different AI PR review bots in parallel, here’s what actually caught real bugs versus what generated noise.

My Team Vibe-Coded a Feature. Three Weeks Later, I'm Still Cleaning Up

Three weeks ago, one of my engineers shipped a feature in two hours that would’ve taken a day. I praised him in the standup. Last week we spent four days unraveling the cascading issues it created. Here’s what vibe coding actually costs a team — and how I’ve changed our process since.

Block Cut ~4,000 Roles and Cited AI Efficiency — What Leaders Should Do Next

Jack Dorsey cut Block’s workforce from 10,000 to ~6,000, citing AI efficiencies. Stock surged 24%. Here’s what’s really happening, what’s overblown, and a concrete playbook for tech leaders navigating AI-driven workforce changes.

Claude Code Review Checklist (2026): 12 Checks Before You Approve

I use this code review checklist for tech leads to reduce false confidence and catch production-grade bugs: idempotency, timeout propagation, auth boundaries, and unsafe retries.

I Ran 5 AI Code Review Tools on the Same PR. Only 2 Were Useful

I ran the same 612-line TypeScript PR through five AI review agents to see which ones actually reduce reviewer workload. Two were consistently useful, two were mostly noisy, and one missed the highest-risk issue entirely. If you’re a tech lead choosing review tooling, this is the practical tradeoff breakdown.

30 Fake AI Extensions Were Stealing My Team's Data — How We Caught It

LayerX researchers found 30+ malicious Chrome extensions impersonating AI assistants, targeting 260,000+ users. One was installed on a developer’s machine on my team. Here’s what happened, what they were stealing, and the 10-minute audit you should run today.

I Tried Figma's Code to Canvas With My Team — Here's What Actually Changed

Figma just launched Code to Canvas with Anthropic — it converts AI-generated code into editable Figma designs. I tested it on three real projects with my team. The design-dev handoff improved, but there are real cracks beneath the polish.

I Let GitHub's Agentic Workflows Run My Repo for a Week. Here's What Happened

GitHub’s new Agentic Workflows let coding agents run inside Actions to triage issues, update docs, and propose fixes — automatically. I ran them on a production repo for a week. Some results were genuinely impressive. Others reminded me why ‘autonomous’ still needs human oversight.

5 Ways I Use AI as an Engineering Manager (That Aren't Coding)

AI isn’t just for writing code. Here’s how I use it for 1:1s, planning, documentation, and the other 60% of a tech lead’s job.

I Spent 3 Hours Configuring MCP Servers Across 4 Editors. Then I Found add-mcp

Every AI coding tool stores MCP server config differently — Claude Code, Cursor, VS Code, Codex all use different files and formats. I wasted hours on manual setup before discovering add-mcp, a CLI that configures MCP servers across all your tools with a single command. Here’s why MCP fragmentation is the biggest DX problem in AI coding right now, and how the ecosystem is starting to fix it.