---
id: "concept-agentic-separation-of-concerns"
type: "concept"
source_timestamps: ["00:07:35", "00:08:40"]
tags: ["system-architecture", "multi-agent-systems", "best-practices"]
related: ["concept-markdown-as-agent-os"]
definition: "The architectural principle that in a multi-agent system, each agent must have a strictly separated identity, toolset, and context to function reliably."
sources: ["s08-real-problem-agents"]
sourceVaultSlug: "s08-real-problem-agents"
originDay: 8
---
# Agentic Separation of Concerns

## Definition

The architectural principle that in a multi-agent system, each agent must have a strictly separated identity, toolset, and context to function reliably.

## The principle

To build a successful multi-agent system (e.g., a marketing-manager agent, a scheduler agent, a CEO agent), there must be strict 'separation of concerns' in the engineering sense. **You cannot have a single 'do-everything' bot.**

Each specialized agent must have:
- Its own distinct identity (own [[concept-markdown-as-agent-os|markdown OS]])
- Its own set of configuration files
- Its own specific toolsets
- Its own workspace
- Clear jurisdictions — they should not share context by default

## What success looks like

Successful implementations resemble *co-workers interacting in Slack*: an orchestrator agent routes a specific question to a specialist agent, who responds within their narrow domain.

## What failure looks like

Without separation, you get **context collapse** — agents step on each other, hallucinate authority, or merge unrelated information. This is the architecturally-correct counterpart to the [[concept-nesting-dolls-management|Nesting Dolls anti-pattern]]: build *specialized* agents with clean boundaries, not stacked auditors patching a confused worker.

## Adjacent literature

Claims-processing literature (V7 Labs) reports ~33% throughput gains from specialized agent pipelines (intake → validation → fraud specialist) — direct support for this principle outside the OpenClaw context.

## Related
- [[claim-generic-agents-are-liabilities]]
- [[framework-markdown-agent-os-architecture]]