automation

Strip away the hype and an AI agent is six plain parts - a model, prompts, skills, tools, access, and memory. The whole mental model, minus the marketing.

The Surprisingly Simple Anatomy of an AI Agent

How Netlify Agent Executor turns Claude, Gemini, and Codex into repeatable multi-step workflows that can review, cross-check, and ship real work.

NAX: Multi-Model Agent Workflows on Netlify

A practical workflow for running multiple coding agents on one branch without duplicate work, file collisions, or constant worktree management.

Multi-Agent Coding Without Worktree Chaos

How to make serverless monorepo deploys faster and safer by deriving deployment waves from git changes, package dependencies, and infrastructure-as-code graphs.

Smart Deploys for Serverless Monorepos

git-er-done turns git diffs into a JavaScript API for smarter CI, monorepo automation, release tooling, and build skips.

David Wells

Builder of things

Software architect
& product wrangler

Get the next post

automation

The Surprisingly Simple Anatomy of an AI Agent

NAX: Multi-Model Agent Workflows on Netlify

Multi-Agent Coding Without Worktree Chaos

Smart Deploys for Serverless Monorepos

git-er-done: Query Git History from JavaScript

David Wells

Builder of things

Software architect& product wrangler

Get the next post

The Surprisingly Simple Anatomy of an AI Agent

NAX: Multi-Model Agent Workflows on Netlify

Multi-Agent Coding Without Worktree Chaos

Smart Deploys for Serverless Monorepos

git-er-done: Query Git History from JavaScript

Software architect
& product wrangler