Five mistakes people make with agentic factories: micromanaging, no quality bar, and 'puppet shows'
hugobowne · x · 2026-09-25
Hugo Bowne teases a free Factory Life lesson with Eleanor Berger on the five mistakes people make when building agentic factories:
- Micromanaging: handing an agent a task, then interrupting constantly to dictate each step — leaving no room to work
- No guidance or quality bar: agents need context, examples, and ways to check their work while keeping room to plan and adapt
- Using a frontier model for everything, regardless of task complexity or cost
- Treating prompts as disposable text instead of infrastructure to maintain
- "The puppet show": inventing a cast of agent personas before knowing which jobs actually need separating
The core question: where do you step in, and where do you let agents get on with it?
More from coding & agent
- Stream launches Agent Skills so coding agents can build chat, video and moderation apps from one prompt — amos_gyamfi · 2026-09-25
- NVIDIA Open-Sources Nemotron-Terminal: Boosts Qwen3-32B from 3.4% to 27.4% on Terminal-Bench 2.0 — _weiping · 2026-09-25
- 5 Ways to Handle Conflicting Information in AI Systems, From Source Ranking to Surfacing Disagreement — goyalshaliniuk · 2026-09-25
- TypeLLM adds type-safe, schema-guaranteed generation to LLMs without touching weights — kalyan_kpl · 2026-09-25
- kuberdenis talks harness engineering and dev workflows with alxfazio, visuals made by Opus 5.5 — tensorqt · 2026-09-25
- Open Claude Code skills catalog ships 31 testing skills, from metamorphic testing to E2E — doodlestein · 2026-09-25