AI Agents
-
Graph engineering for agents: what a prompt chain forgets
Graph engineering models an agent system as nodes and edges, not a prompt chain. Here is the idea, five patterns, and how my open-source Yaaf uses it.
-
MCP Skills: I Tried Them, and Verified Is Not Trusted
MCP skills ship in the official TypeScript SDK 2.4. I ran them, and a hostile server's prompt injection still came back verified: true. Here is why.
-
Matt Pocock Skills: Every Skill Explained and How to Install
Matt Pocock skills explained: what all 27 plugin skills do, who can call each, how to install, what critics say, and how it compares to GSD and BMAD.
-
WebMCP: I Tested Chrome’s Tool API for Browser Agents
WebMCP lets a web page hand tools to a browser agent. I ran it in Chromium 153, hit one real bug, and sketched a half-day Trellis experiment.
-
Codebase Port Cost: What ts-rust Really Took
ts-rust ported the TypeScript compiler to Rust with agents. I ran it and read its own numbers: the codebase port cost $24k after a $400k stall.
-
The Hidden Risk of Letting AI Manage Ad Campaigns
MCP's own specs disagree on whether it is stateful. Neither defines a record of what changed in your ad account — and an unattended run needs that record.
-
Your Agent A/B Test Was Broken Before You Ran It
In parimutuel betting your own bet moves the odds you get, so you can't test a strategy against half your bankroll. Agents have the same problem, and almost nobody running an A/B test on one has noticed.
-
Half My Agent’s Budget Goes to Checking Its Own Work
I built the quality gate last, in an afternoon, as the cheap safety rail. Then I added up the API bills by phase. Across four runs it took 51.5% of everything the pipeline spent.
-
I Rewrote the Test Until My Writing Passed It
Five commits in four days, every one adjusting the grader that judges this blog's drafts, every one triggered by a verdict I disagreed with. Finance has been making that mistake with real money since the seventies.
-
The Bug Was Reproducible. My Logs Weren’t.
A post on this site ended in a raw API payload for 151 days. Fixing it took two minutes. Explaining it took longer, and I never finished, because I had not recorded what the model actually said.