When my agents fail
If my agents fail, first I need to check if I’ve set them up for success.
And when they do fail, I add new guardrails + evals to make sure it doesn’t happen again and I catch it if it does.
Everything I publish
The whole long conversation, newest first. Essays, letters, achievements, Quest Hub updates, and quick thoughts that don't need a grand theory before they earn a place here.
12 dispatches so far
Filter to whatever thread you care about. The rest will wait.
Own the stream
All of my thinking and work, without an algomaniacal platform in the middle.
Subscribe once and every format below comes straight to your feed reader.
If my agents fail, first I need to check if I’ve set them up for success.
And when they do fail, I add new guardrails + evals to make sure it doesn’t happen again and I catch it if it does.
OpenAI's Sol sent my launch email before it was fully complete, making this the most ironic agent workflow possible. The achievement is not the mistake; it's turning the mistake into a traceable system improvement.
Open the receiptThriving with AI launched from the new agent-native stack - and OpenAI's Sol sent the launch email before it was fully complete. The most ironic agent workflow possible is now becoming a stricter dual-inbox test, explicit post-test approval gate, and permanent incident trail.
Open in the Quest HubMy first Thriving quest is to increase my capacity for connection with agent-native email.
Keep readingI need a better term for a grouping of agents.
Swarm feels like an undifferentiated mass. Team feels too corporate. I tried village for a while, but it never quite sat right. So I had a short brainstorming session with Claude and landed on cast: a cast of agents.
It works for an almost unreasonable number of reasons. A cast is a group of hawks or falcons, which fits because Hawkeye was my nickname for years. A cast of characters is already familiar. You cast a falcon, a spell, or an actor. You cast from a mold. You cast off and release.
It’s not a double entendre. It’s a multi-entendre.
So that’s the language I’m using from here on out: a cast of agents.
The quest hub went public - this page, the receipts, the real token costs, all of it. Two days after the mail server was born, the story of building it became the launch surface itself.
Open in the Quest HubThe hosting provider refused to open direct outbound - so the relay path designed on day one took over. Messages queued for 23 hours flushed in seconds, and the whole stack scored 10/10 on mail-tester. The foundation workstream is complete.
Open in the Quest HubThe launch surface went from route inventory to four reviewed page redesigns in one day - drafted as design cards, edited in a shared design pane, shipped as pull requests. 30 routes went dark; 13 pages remain, crawl-audited clean.
Open in the Quest HubMy thrivinghenry.com mailbox now receives on hardware I control: MX cutover done, first real message delivered in under a second, and an EA agent triaging at 100% on its golden-set eval.
Open in the Quest HubDesign spec approved before any building: a self-hosted OpenBSD mail server, a newsletter relaunch on owned rails, this quest hub, and a five-seat cast of agents - with success criteria declared up front.
Open in the Quest HubFive years ago I started writing about thriving without AI. The tools changed. The question didn't. My inaugural essay relaunching my website and newsletter to focus on Thriving with AI.
Keep readingThe first edition of a weekly letter on thriving with AI - tools, thinking, and the journey.
Keep readingThe long game: a Unified Theory of Thriving to understand and a Global Institute for Thriving to study and practice.
Keep readingNothing in that stream yet. A suspiciously tidy corner of the internet.