Writing

Essays on running AI agents in production, each built on something I measured in my own workspace, and a digest of the public work as it ships. New pieces also go out on Substack, where you can get them by email.

Latest

Essays

No configuration check failed during four days with no output.

29 May 2026 · 9 min read

The self-auditing workspace

An agent workspace runs while you sleep, and rots while you sleep. So mine audits itself about once a week. The hard part wasn't building the audit. It was the three ways a self-audit quietly lies to you.

A two-column table of what Copilot gives you and what you build around it.

27 May 2026 · 10 min read

Copilot is not the workspace

How to put a stateless enterprise LLM assistant inside a workspace that gives it the persistence, provenance, and validation it doesn't have on its own.

Of 49 task bullets, 53% had a default, 35% needed intent and 12% were out of scope.

24 May 2026 · 8 min read

Build first, ask second

Turning a scheduled agent from a question-asker into a sandboxed teammate that ships work for review.

Digest

  1. 16 Nov 20264mDigest #4: three teardowns, a new Action and six experimentsdigest
  2. 27 Sep 20263mDigest #3: a teardown of GPT-RAG and two correctionsdigest
  3. 06 Sep 20262mDigest #2: three teardowns and a learn trackdigest
  4. 28 Aug 20262mDigest #1: three patterns, a linter, and a mapdigest

New pieces go out by email on Substack, and there's an RSS feed.