The field manual · 4 essays
Essays on proactive agents.
Adapted from internal positioning docs, postmortems, and the kind of design notes that usually live and die in PR descriptions. New ones appear roughly monthly — weather permitting.
- N°04
May 4, 2026 · 6 min
Proactive agents need three primitives
An agent isn't proactive because it has a heartbeat. It's proactive because it has a clock, a watcher, and an inbox — wired together with durable state. Here's why the triple is the unit, and what we built around it.
Read the essay
- N°03
April 20, 2026 · 4 min
Reactive vs proactive, with examples
A short tour of how the same agent looks under both architectures. Same goal, same provider, different posture toward the world. The difference is mostly about who waits for whom.
Read the essay
- N°02
April 6, 2026 · 5 min
The eight-week webhook tax
Adding one provider to a proactive agent — the honest, production-ready, post-incident-review way — is at minimum a sprint. Doing it for four providers is most of a quarter. We mapped the cost so you don't have to discover it.
Read the essay
- N°01
March 9, 2026 · 3 min
Why we stopped saying multi-agent coordination
We had the wrong elephant. Multi-agent coordination is real, but it's a property of a runtime, not a category to sell. Here's what changed when we stopped leading with it.
Read the essay