Problem #5 of running 100 AI agents: observability, security and per-agent cost tracking
goyalshaliniuk · x · 2026-11-07
Part 5 of the series: with 100 agents running simultaneously, it becomes hard to answer which agent caused a failure, which workflow consumes the most tokens, whether agents access tools they shouldn't, and how to trace a task across agents. Remedies: distributed tracing, per-agent cost tracking, centralized logs, access controls, and continuous evaluations. The post recommends Langfuse, which claims 21 of the Fortune 50 as users, 90B+ monthly observations, and 100k+ engineers building on it.
Related event: Scaling from 1 to 100 AI Agents: Five Engineering Pillars You Need First(7 posts)→
More from coding & agent
- Claude auto-renders 21k Blender images to train a model that recreates your LEGO build — nptacek · 2026-11-07
- The anti-slop AI coding prompt that works: "deepen modules, simplify interfaces" — mattpocockuk · 2026-11-07
- AI coding widens the engineer gap: average users vs power users is night and day — bendee983 · 2026-11-07
- Matt Pocock: to constrain coding agents to happy paths, you need to know the map — mattpocockuk · 2026-11-07
- Survey of 520 tech leaders: only 8% use one model provider, 80%+ delayed migrations over switching costs — ombre-ai · 2026-11-07
- Claude paints watercolor trees via custom-SDK abstractions, not direct robot control — teortaxesTex · 2026-11-07