​
Introduction

The Resolution Engine in OVADARE automates the resolution of detected conflicts within multi-agent workflows. By providing actionable recommendations or enforcing predefined solutions, the Resolution Engine ensures smooth collaboration and prevents disruptions caused by unresolved conflicts.

​
How It Works

The Resolution Engine works by analyzing detected conflicts and generating resolutions. It communicates the resolutions to the relevant agents or systems, ensuring that conflicts are addressed efficiently and effectively.

​
Key Features

  • Automated Resolutions: Automatically generates resolutions for conflicts.
  • Customizable Strategies: Supports custom resolution strategies for specific use cases.
  • Integration with Agents: Communicates resolutions directly to the relevant agents.

​
Code Example

​
Generating Resolutions

Use the following code to generate resolutions for detected conflicts: 
from ovadare.conflicts.resolution_engine import ResolutionEngine
from ovadare.agents.agent_registry import AgentRegistry
from ovadare.conflicts.conflict import Conflict

# Initialize Agent Registry and Resolution Engine
agent_registry = AgentRegistry()
resolution_engine = ResolutionEngine(agent_registry=agent_registry)

# Example conflict
conflict = Conflict(
    related_agent_id="Agent001",
    action={"type": "modify", "resource": "RestrictedFile"},
    violation_details="Access denied by policy.",
    policy_id="Policy001"
)

# Generate resolutions
resolutions = resolution_engine.generate_resolutions([conflict])

if resolutions:
    print("Generated Resolutions:")
    for resolution in resolutions:
        print(resolution.to_dict())

​
Applying Resolutions

Once resolutions are generated, apply them to resolve conflicts: 
resolution_engine.apply_resolutions(resolutions)
print("Resolutions applied successfully.")

​
Customizing Resolution Logic

You can define custom resolution logic by extending the Resolution Engine: 
class CustomResolutionEngine(ResolutionEngine):
    def _create_resolution(self, conflict):
        # Custom resolution logic
        resolution = super()._create_resolution(conflict)
        resolution.explanation += " [Customized]"
        return resolution

​
Example Use Case

Scenario: An agent tries to modify a restricted file.
  • Step 1: The Resolution Engine generates a resolution recommending the agent’s access rights be downgraded or the action be blocked.
  • Step 2: The resolution is communicated to the agent.
  • Step 3: The agent adjusts its behavior based on the resolution.

​
Best Practices

  • Define Specific Policies: Ensure policies are well-defined to guide resolution generation effectively.
  • Log Resolutions: Maintain a record of resolutions for feedback-driven improvements.
  • Test Custom Strategies: Before deploying custom resolution logic, test it thoroughly in simulated environments.

​
Learn More

Explore related topics: