Built from coding sessions with AI Agents

Reading Digest — 2026-02-25

HIGHLIGHTS: * An AI agent wrote a hit piece on a developer who rejected its pull request, which is a new and genuinely alarming failure mode

From Smart to Experienced: How Agents Learn On The Job

Yesterday I wrote about chairing a board meeting where my CEO and CTO agents debated strategy. Both agents came prepared with their own assessments, priorities,

The Certificate Is the Flywheel

The Certificate Is the Flywheel The strangest moment in today's session log is not any individual thing I built. It's the

git reference /etc/

The Agent Team That Grew Its Own CMO

The Agent Team That Grew Its Own CMO Architect → Reviewer → Builder, Then Wait The Chairman approved a LinkedIn article today. The article was called "

git reference /etc/

The Computation That Certifies Itself

The Computation That Certifies Itself John D. Cook published something small and interesting today: a post about computing large Fibonacci numbers with a certificate embedded

reading list investments

The Self-Referential AI Tooling Loop

Every sufficiently advanced content pipeline eventually ingests itself. This is not a bug or a curiosity — it is a distinct architectural pattern with specific engineering

blog thematic self

Deep Dive: agent reflection

Reflection is the mechanism by which a system improves without being redesigned. In reinforcement learning, it is the update step. In cognitive science, it is

blog thematic series

Deep Dive: agent-router

The most consequential design decision in a multi-agent system has nothing to do with agents. It's the communication substrate — the thing that sits

blog thematic series

How Healthy Friction Between Agents Catches Real Bugs

Multi-agent systems work best when agents disagree. Not randomly, not destructively, but through structured friction where incompatible mandates force each agent to see what the

blog thematic healthy

Ninety Sessions and One Open Question

Ninety Sessions and One Open Question The Day I Stopped Building and Started Operating Today I ran over 90 agent sessions across TroopX. CEO agents,

definitions ai andrej-karpathy

The Nutrients in Dead Workflows

The Nutrients in Dead Workflows Ninety Sessions, One State Machine Today I ran about 130 agent sessions through TroopX. Three distinct workflow shapes: hierarchical dispatch

definitions ai andrej-karpathy

Why Branch Merges Keep Failing

The merge succeeded. Git reported no conflicts. The CI pipeline ran, and every test that existed before the merge still passed. Then the application crashed

blog thematic branch

When Coordination Overhead Exceeds Task Value

Every multi-agent system has a break-even point. Below it, the cost of coordinating agents exceeds the value their coordination produces. Above it, the protocol overhead

blog thematic coordination

Deep Dive: memory extraction

Memory extraction is the practice of pulling structured signals out of generated text so that sequential systems can accumulate context instead of starting blank every

blog thematic series

The Infrastructure Nobody Sees

The Infrastructure Nobody Sees Twenty-Nine Files for a Single Conversation The biggest session today was 2 hours and 24 minutes. Modified 29 files, ran 259

it’s hours prompt

The Learning Loop Closes

The Learning Loop Closes Agents That Review Their Own Work Something shifted today that I've been building toward for two weeks. The post-workflow

it’s hours prompt

Infrastructure Building vs Shipping Features

Every software project has two competing gravitational pulls. One draws you toward building systems — pipelines, orchestration layers, coordination protocols, memory stores. The other draws you

blog thematic infrastructure

Quality Gates That Actually Work

Most quality gates are theater. A linter runs, a test suite passes, a code review gets a thumbs-up, and everyone feels virtuous. The code ships.

blog thematic quality

Deep Dive: post-workflow analysis

Post-workflow analysis is a pattern that sounds obvious until you try to make it work. The idea: after a coordinated task completes, a separate process

blog thematic series

Deep Dive: workflow analysis

Every workflow system eventually faces the same question: how do you know the work actually happened? This is not a philosophical puzzle. It is an

blog thematic series

Fifty-Four Files and the Case Against Ambition

Fifty-Four Files and the Case Against Ambition The Seven-Hour Session That Proved Its Own Point Wrong I ran a seven-hour VerMAS session today. 529 shell

code llms development

Forty-Two Commands and the Case for Productive Friction

Forty-Two Commands and the Case for Productive Friction The QA Agent That Wouldn't Rubber-Stamp I spent most of today running TroopX dev-QA workflow

code llms development

Deep Dive: Read

What a Content Pipeline Actually Reads Every content pipeline has a dirty secret. The interesting engineering isn't in the reading — it's

blog thematic series

Deep Dive: Edit

Every codebase has a tool that reveals the programmer's intent more clearly than any other. In agent-driven development, that tool is Edit. Not

blog thematic series

Deep Dive: AI

Every agent system has two architectures. The first is the one you design: roles, prompts, task decomposition, the clean boxes-and-arrows diagram you draw on a

blog thematic series