Discussion Pattern - TFrameX Pattern Example

Demonstrates collaborative agent discussions where multiple agents engage in iterative dialogue to reach consensus, explore different perspectives, or solve complex problems through multi-agent conversation.

🎯 What You'll Learn

Multi-Agent Dialogue: Facilitating conversations between multiple agents
Perspective Synthesis: Combining different viewpoints into unified insights
Consensus Building: Reaching agreement through structured discussion
Discussion Moderation: Guiding productive multi-agent conversations
Collaborative Problem-Solving: Leveraging diverse expertise through dialogue

📁 Project Structure

discussion-pattern/
├── README.md              # This guide
├── requirements.txt       # Dependencies
├── .env.example          # Environment template
├── main.py               # Main application
└── docs/
    └── discussion_flows.md # Advanced discussion patterns

🚀 Quick Start

# Install dependencies
pip install -r requirements.txt

# Configure environment
cp .env.example .env
# Edit .env with your LLM settings

# Run the example
python main.py

🗣️ Discussion Pattern Architecture

                    Discussion Topic
                          │
                          ▼
                 DiscussionModerator
                    (Facilitator)
                          │
        ┌─────────────────┼─────────────────┐
        ▼                 ▼                 ▼
   OptimistAgent    RealistAgent    SkepticalAgent
   (Opportunities)   (Practical)     (Risks)
        │                 │                 │
        └─────────────────┼─────────────────┘
                          ▼
                  CreativeAgent
                   (Innovation)
                          │
                          ▼
                 ConsensusBuilder
                  (Synthesis)
                          │
                          ▼
                 Final Agreement

Collaborative Discussion Flow

Moderation: DiscussionModerator sets agenda and guides process
Multi-Perspective Input: Various agents share diverse viewpoints
Iterative Dialogue: Multiple rounds of discussion and refinement
Synthesis: ConsensusBuilder integrates perspectives into unified outcomes

💻 Example Usage

Business Strategy Discussion

# Topic: "Should we invest in AI automation for customer service?"
#
# Round 1 - Initial Perspectives:
# → OptimistAgent: "Great opportunity for efficiency and 24/7 support!"
# → RealistAgent: "Need to consider implementation costs and training"
# → SkepticalAgent: "Risk of losing personal touch with customers"
# → CreativeAgent: "What about hybrid AI-human approach?"
#
# Round 2 - Synthesis:
# → ConsensusBuilder: "Phased implementation with human oversight..."
# → Final Decision: Balanced approach addressing all concerns

🎮 Demo Modes

1. Business Strategy Discussion

Complete strategic decision-making discussion:

python main.py
# Select option 1

2. Product Development Discussion

Feature prioritization through collaborative dialogue:

python main.py
# Select option 2

3. Crisis Management Discussion

Multi-round crisis response planning:

python main.py
# Select option 3

4. Individual Perspective Analysis

See how different agents approach the same question:

python main.py
# Select option 4

5. Consensus Building Demo

Watch consensus emerge from diverse viewpoints:

python main.py
# Select option 5

6. Interactive Discussion Mode

Join the discussion as a participant:

python main.py
# Select option 6

📊 Discussion Pattern Benefits

🎯 Comprehensive Analysis

Multiple expert perspectives
Thorough exploration of options
Reduced blind spots and bias
More robust decision-making

🤝 Consensus Building

Structured path to agreement
Address all stakeholder concerns
Build buy-in and commitment
Create win-win solutions

💡 Creative Problem-Solving

Diverse thinking approaches
Cross-pollination of ideas
Breakthrough insights
Innovation through collaboration

⚖️ Balanced Outcomes

Consider multiple viewpoints
Weigh pros and cons thoroughly
Mitigate risks while pursuing opportunities
Make well-rounded decisions

🏗️ Building Discussion Systems

1. Define Discussion Roles

@app.agent(
    name="ModeratorAgent",
    description="Facilitates productive discussion",
    system_prompt="Guide conversation, ensure participation, drive consensus..."
)
async def moderator_agent():
    pass

@app.agent(
    name="PerspectiveAgent",
    description="Represents specific viewpoint", 
    system_prompt="Focus on [specific perspective]: opportunities, risks, practicality..."
)
async def perspective_agent():
    pass

2. Create Discussion Flow

async def facilitate_discussion(topic: str, rounds: int = 3):
    async with app.run_context() as rt:
        for round_num in range(rounds):
            # Moderator guides
            moderator_result = await rt.call_agent("ModeratorAgent", topic)
            
            # Participants contribute
            for participant in participants:
                result = await rt.call_agent(participant, topic)
                
            # Synthesize round
            synthesis = await rt.call_agent("ConsensusBuilder", topic)

3. Manage Discussion Flow

class DiscussionManager:
    def __init__(self, participants: List[str]):
        self.participants = participants
        self.discussion_history = []
    
    async def run_discussion(self, topic: str) -> str:
        # Implement structured discussion logic
        pass

🎯 Use Cases

🏢 Strategic Planning

Business strategy decisions
Investment evaluations
Market entry planning
Resource allocation

🚀 Product Development

Feature prioritization
Design decisions
Technology choices
User experience planning

⚖️ Policy Making

Regulatory compliance
Ethics considerations
Risk management
Stakeholder alignment

🔄 Process Improvement

Workflow optimization
Quality enhancement
Efficiency initiatives
Change management

🔧 Advanced Discussion Patterns

Structured Debate Format

@app.agent(
    name="DebateModerator",
    system_prompt="""
    Moderate a structured debate:
    1. Opening statements from each side
    2. Argument exchange rounds
    3. Rebuttal opportunities  
    4. Closing arguments
    5. Final synthesis
    """
)

Devil's Advocate Pattern

@app.agent(
    name="DevilsAdvocate", 
    system_prompt="""
    Challenge the emerging consensus by:
    - Questioning assumptions
    - Identifying weaknesses
    - Proposing alternative views
    - Stress-testing decisions
    """
)

Socratic Method Discussion

@app.agent(
    name="SocraticModerator",
    system_prompt="""
    Guide discussion through questions:
    - Ask probing questions
    - Challenge assumptions
    - Draw out implications
    - Lead to insights through inquiry
    """
)

📈 Discussion Optimization

Participant Balance

Ensure diverse perspectives
Avoid dominant voices
Encourage equal participation
Manage group dynamics

Flow Management

Set clear discussion structure
Monitor progress toward goals
Keep conversations focused
Prevent endless loops

Quality Control

Ensure substantive contributions
Prevent superficial responses
Maintain professional dialogue
Drive toward actionable outcomes

🔍 Discussion Evaluation

Consensus Quality Metrics

def evaluate_consensus(discussion_output: str) -> dict:
    return {
        "agreement_level": assess_agreement(discussion_output),
        "concern_coverage": check_concerns_addressed(discussion_output),
        "actionability": evaluate_actionable_outcomes(discussion_output),
        "stakeholder_satisfaction": rate_stakeholder_buy_in(discussion_output)
    }

Participation Analysis

def analyze_participation(discussion_log: List[dict]) -> dict:
    return {
        "contribution_balance": measure_equal_participation(discussion_log),
        "perspective_diversity": assess_viewpoint_variety(discussion_log),
        "quality_of_input": rate_contribution_substance(discussion_log)
    }

Decision Quality Assessment

Comprehensiveness of analysis
Stakeholder concern integration
Implementation feasibility
Risk mitigation effectiveness

🚀 What's Next?

After mastering discussion patterns:

Advanced Integration: Content Creation Pipeline
Real-World Applications: Legal Document Analysis
Complex Workflows: DevOps Incident Response
Build Custom Discussions: Create your own collaborative agent systems

💡 Best Practices

Design Principles

Clear Roles: Define distinct perspectives and responsibilities
Structured Process: Use consistent discussion frameworks
Balanced Participation: Ensure all voices are heard
Outcome Focus: Drive toward actionable decisions

Implementation Tips

Start with 3-4 distinct perspectives
Use a strong moderator to guide flow
Build in synthesis and consensus steps
Test with various topic complexities

Common Pitfalls

Avoid echo chambers (too similar perspectives)
Don't let discussions become endless
Ensure moderator remains neutral
Plan for when consensus isn't reached

📚 Further Reading

📄 License

This example is provided under the MIT License.

Uh oh!

FilesExpand file tree

README.md

Latest commit

History