updates m6

Author: pablorodMS

learn-pr/wwl/design-extensibility-ai-solutions/includes/3-design-agents-microsoft-365-copilot.md

@@ -172,48 +172,6 @@ Will the agent support frontline or knowledge workers?
 
 What latency, safety, or reliability constraints apply?
 
-## Professional visuals (text-based for Word)
-
-### Visual 1 - Agent architecture model
-
-User Action 
-
-    ↓
-
-Microsoft 365 Application (Teams, Outlook, Loop)
-
-    ↓
-
-Copilot Agent Instructions & Policies
-
-    ↓
-
-Tools & Connectors (Graph, SharePoint, APIs)
-
-    ↓
-
-Business Logic Execution
-
-    ↓
-
-Response Back to User
-
-### Visual 2 - Collaborative agent workflow
-
-Primary Agent (Intent Detection)
-
-      ↓ determines task
-
-SubAgent A (Data Gathering)    SubAgent B (Draft Generation)
-
-      ↓                                 ↓
-
-     Combined Output → Human Review → Publish/Act
-
-### Visual 3 - Agent governance lifecycle
-
-Design → Build → Test → Deploy → Monitor → Improve
-
 ## References
 
 [https://learn.microsoft.com/en-us/copilot/microsoft-365/microsoft-365-copilot-page?view=o365-worldwide](/copilot/microsoft-365/microsoft-365-copilot-page)

learn-pr/wwl/design-extensibility-ai-solutions/includes/6-design-agents-automate-tasks-apps-websites-computer-copilot-studio.md

@@ -216,20 +216,6 @@ Build monitoring and fallback plans
 
 Use Computer Use only when APIs are unavailable
 
-## Professional visuals (text-based)
-
-### Visual 1 — Computer Use architecture
-
-User Request ↓ Agent Reasoning ↓ Computer Use Actions ↓ UI Interaction ↓ Validation & Response 
-
-### Visual 2 — When to Use Computer Use
-
-No API Available | UI-Driven Tasks | Multi-App Workflows | Human-Like Reasoning Needed 
-
-### Visual 3 — Agent design framework
-
-Define Task → Add Context → Break Steps → Add Guardrails → Test → Refine 
-
 ## References
 
 [https://learn.microsoft.com/en-us/microsoft-copilot-studio/computer-use](/microsoft-copilot-studio/computer-use) 

learn-pr/wwl/design-extensibility-ai-solutions/includes/7-design-agent-behaviors-copilot-studio-reasoning-voice-mode.md

@@ -162,66 +162,6 @@ Agent provides concise spoken output
 
 Because voice relies on real-time interaction, architects should design behavior models that prioritize responsiveness, fallback handling, and user-friendly error messages.
 
-**Architectural model for agent behaviors**
-
-## Text-based visual: Agent behavior architecture
-
-### Visual A — Agent Behavior Stack
-
-+------------------------------------------------------+
-
-| User Interaction Layer (Text, Voice, UI Controls)    |
-
-+------------------------------------------------------+
-
-| Dialogue Management & Conversation Flow              |
-
-+------------------------------------------------------+
-
-| Reasoning Engine (Standard or Deep Reasoning)        |
-
-+------------------------------------------------------+
-
-| Instruction Layer (Rules, Roles, Tone, Guardrails)   |
-
-+------------------------------------------------------+
-
-| Knowledge & Data Access (SharePoint, Dataverse, etc.)|
-
-+------------------------------------------------------+
-
-| Tools and Actions (APIs, Connectors, Automations)    |
-
-+------------------------------------------------------+
-
-| Governance & Safety (Policies, Filters, Logging)     |
-
-+------------------------------------------------------+
-
-**Design patterns for high-quality agent behavior**
-
-## Key behavior patterns
-
-### Pattern 1 — Step-by-step reasoning
-
-Break tasks into steps to improve accuracy.
-
-### Pattern 2 — Validation Prompting
-
-Ask the agent to verify assumptions or confirm missing context.
-
-### Pattern 3 — Context Reuse
-
-Carry summaries and reasoning from earlier messages in the same session.
-
-### Pattern 4 — Intent Routing
-
-Identify user purpose and route to knowledge, actions, or workflows.
-
-### Pattern 5 — Behavioral Guardrails
-
-Prevent overreach by defining what the agent must never do.
-
 ## References
 
 [https://learn.microsoft.com/en-us/microsoft-copilot-studio/guidance/agents/agent-behaviors](/microsoft-copilot-studio/guidance/agents/agent-behaviors)