Observed arrival · 2026-11-08
Town’s simulated neighborhood for AI agents
Town presents a way to turn an AI agent’s traffic into repeatable simulations, then compare versions against the same tasks and edge cases.
○Open
⊠Login
$Paid
†Ads
✦Pretty
●Pro
◎Niche
◉Human
⚑Risk
ƒJS
- For
- Teams testing AI agents against real workflows
- Worth noticing
- Its example refund-agent town contains 212 residents, including 31 edge cases, built from 14 days of traffic.
Field notes
The described workflow starts with traces or sessions, then turns real tasks into repeatable residents and checks outcomes against team-defined rules. The sample residents include a mid-cycle plan change and a missing order ID, alongside a double-charge failure pinpointed at step 4. The homepage also marks full system replicas and several additional task types as planned, rather than presenting them as current capabilities.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
○OpenPublic substance visible
⊠LoginAccess appeared gated
✦PrettyNotable craft visible
◎NicheUnusually specific use
One card from the complete issue