The Company Selling You AI Product Managers Doesn't Give Its Own Agents Job Titles: Roles, Isolation, and Code in Multi-Agent Systems

卖你 AI 产品经理的公司，自己的 agent 不设岗位：multi-agent 的角色、隔离与代码

Grok Bot markets agents as named coworkers like sales or finance, but its engineering core Grok Build uses only functional names such as researcher-0 and verifier—zero personas. The article breaks multi-agent design into three layers: the mechanism layer relies on context-window isolation for output quality; the orchestration layer decides who controls the next step (teammate persona, main agent, or script); the interface layer uses job hats to lower the human adoption barrier. Hats solve three human problems—delegation intuition, approval anchors, and memory partitioning—but do nothing for model reasoning. All coworkers under one account share the same cloud computer and credentials; security boundaries depend solely on manual approval gates. When building your own system, nail window isolation first, then pick an orchestration style based on task reliability needs, and save persona packaging for last.

Why it matters: Hits all three HKR axes. Strong headline hook, concrete product logs backing the three-layer breakdown, directly addresses a daily pain point for agent builders. Docked slightly because it's a solo blog post without cross-source corroboration, but the analytical framework itse...