Conflict Monitor Plugin

Package: @heretek-ai/conflict-monitor-plugin
Version: 1.0.0
Type: ACC Brain Function Implementation
License: MIT

Overview

The Conflict Monitor Plugin implements Anterior Cingulate Cortex (ACC) functions for the Heretek OpenClaw collective. It provides real-time conflict detection, severity scoring, and resolution suggestions for agent proposals and goals.

Brain Function Mapping

| Brain Region | Function | Implementation | |--------------|----------|----------------| | Anterior Cingulate Cortex (ACC) | Conflict monitoring | ConflictDetector class | | ACC | Error detection | Severity scoring with escalation | | ACC | Cognitive control | Resolution suggestion generation | | Prefrontal Cortex | Decision support | Triad deliberation integration |

Features

• Real-time Conflict Detection - Monitors agent proposals for logical contradictions, goal conflicts, resource contention, value violations, and temporal conflicts
• Severity Scoring - Multi-factor assessment with low/medium/high/critical levels
• Resolution Suggestions - Strategy-based recommendations (compromise, collaboration, arbitration, etc.)
• History Tracking - Complete conflict history with analytics
• Triad Integration - Direct integration with triad deliberation protocol

Installation

cd plugins/conflict-monitor
npm install

Quick Start

import { createPlugin } from '@heretek-ai/conflict-monitor-plugin';

// Initialize plugin
const plugin = await createPlugin({
  triadIntegration: true,
  autoDetectConflicts: true,
  autoGenerateSuggestions: true
});

// Register agents
plugin.registerAgent('alpha', {
  goals: ['Optimize reasoning efficiency'],
  proposals: []
});

// Analyze a proposal
const result = await plugin.analyzeProposal({
  id: 'proposal-1',
  agentId: 'beta',
  content: 'We should prioritize speed over accuracy',
  goals: ['Complete tasks quickly']
});

console.log(`Detected ${result.conflicts.length} conflicts`);
console.log(`Highest severity: ${result.summary.highestSeverity}`);

Conflict Detection

Conflict Types

| Type | Description | Example | |------|-------------|---------| | logical_contradiction | Direct logical inconsistency | "Enable X" vs "Disable X" | | goal_conflict | Incompatible objectives | "Maximize speed" vs "Ensure thoroughness" | | resource_conflict | Competition for resources | Two agents needing exclusive CPU access | | value_conflict | Value system violations | "Autonomy" vs "Control" | | temporal_conflict | Scheduling overlaps | Same time slot for two tasks | | authority_conflict | Jurisdiction disputes | Two agents claiming same authority | | methodology_conflict | Approach disagreements | Different implementation strategies |

Detection Algorithms

The plugin uses multiple detection algorithms:

1. Negation Detection - Identifies direct "A" vs "not-A" contradictions
2. Mutual Exclusivity - Detects inherently incompatible goals
3. Pattern Matching - Matches against known contradiction patterns
4. Resource Analysis - Checks for exclusive resource requirements
5. Value Opposition - Identifies opposing values in the value system
6. Temporal Overlap - Calculates time slot conflicts

Severity Scoring

Scoring Factors

| Factor | Weight | Description | |--------|--------|-------------| | Autonomy Impact | 15% | Impact on agent autonomy | | Collective Impact | 20% | Impact on collective objectives | | Agent Count | 10% | Number of agents affected | | Resource Contention | 15% | Level of resource competition | | Value Violation | 20% | Severity of value violations | | Temporal Urgency | 10% | Time sensitivity | | Escalation Potential | 10% | Risk of conflict escalation |

Severity Levels

| Level | Score Range | Description | |-------|-------------|-------------| | low | 0.0 - 0.3 | Minor conflicts, log only | | medium | 0.3 - 0.6 | Moderate conflicts, monitor | | high | 0.6 - 0.85 | Serious conflicts, intervention needed | | critical | 0.85 - 1.0 | Emergency, immediate action required |

Resolution Strategies

| Strategy | Description | Success Rate | Use Case | |----------|-------------|--------------|----------| | compromise | Find middle ground | 65% | Most conflicts | | collaboration | Win-win solution | 55% | High-trust situations | | accommodation | One party yields | 70% | Low-priority conflicts | | competition | Winner takes all | 50% | Clear merit cases | | avoidance | Delay resolution | 40% | Low-urgency conflicts | | split_difference | Equal division | 60% | Resource conflicts | | arbitration | Third-party decision | 75% | High/critical severity | | consensus | Everyone agrees | 50% | Triad deliberations | | reframing | New perspective | 45% | Value conflicts | | resource_expansion | Expand resources | 65% | Resource scarcity |

API Reference

Class: ConflictMonitorPlugin

Constructor Options

{
  // Detection settings
  sensitivity: 0.7,
  enableLogicalDetection: true,
  enableGoalDetection: true,
  enableResourceDetection: true,
  enableValueDetection: true,
  enableTemporalDetection: true,
  knownContradictions: [],
  valueSystem: [],
  
  // Scoring settings
  factorWeights: {},
  severityThresholds: {},
  contextMultipliers: {},
  criticalEscalationThreshold: 0.85,
  autoEscalate: true,
  agentPriorities: {},
  
  // Resolution settings
  enabledStrategies: [],
  minSuccessRate: 0.3,
  maxSuggestions: 5,
  includeSteps: true,
  useHistoricalData: true,
  
  // Plugin settings
  triadIntegration: true,
  triadMembers: ['alpha', 'beta', 'charlie'],
  autoDetectConflicts: true,
  autoGenerateSuggestions: true,
  notifyOnCritical: true,
  enableAnalytics: true,
  analyticsInterval: 60000,
  maxHistorySize: 1000
}

Methods

initialize(options)

Initialize the plugin.

await plugin.initialize({
  triadIntegration: true,
  autoGenerateSuggestions: true
});

registerAgent(agentId, state)

Register an agent for monitoring.

plugin.registerAgent('alpha', {
  goals: ['Goal 1'],
  proposals: [],
  resources: [],
  values: []
});

analyzeProposal(proposal, options)

Analyze a proposal for conflicts.

const result = await plugin.analyzeProposal({
  id: 'proposal-1',
  agentId: 'alpha',
  content: 'Proposal content',
  goals: ['Goal 1', 'Goal 2']
}, {
  context: {
    isTriadDeliberation: true,
    urgency: 'high'
  }
});

Returns:

{
  proposalId: 'proposal-1',
  conflicts: [...],
  severities: [...],
  suggestions: [...],
  summary: {
    totalConflicts: 2,
    severityCounts: { low: 1, high: 1 },
    highestSeverity: 'high',
    requiresAttention: true
  }
}

monitorTriadDeliberation(deliberation)

Monitor triad deliberation for conflicts.

const result = await plugin.monitorTriadDeliberation({
  id: 'deliberation-1',
  phase: 'voting',
  participants: ['alpha', 'beta', 'charlie'],
  proposals: [...]
});

resolveConflict(conflictId, resolution)

Mark a conflict as resolved.

plugin.resolveConflict('conflict-123', {
  strategy: 'compromise',
  description: 'Both parties agreed to split resources',
  success: true,
  resolvedAt: Date.now()
});

getSuggestions(conflictId, options)

Get resolution suggestions for a conflict.

const suggestions = plugin.getSuggestions('conflict-123');

getAnalytics()

Get comprehensive analytics.

const analytics = plugin.getAnalytics();
// { conflicts, severity, resolutions, triadStatus }

Events

| Event | Payload | Description | |-------|---------|-------------| | initialized | { name, version, triadIntegration } | Plugin initialized | | conflictDetected | ConflictDetectionResult | New conflict detected | | severityAssessed | { conflict, severity } | Severity assessed | | criticalConflict | { conflict, severity, suggestions } | Critical conflict detected | | conflictResolved | { conflictId, resolution } | Conflict resolved | | analyticsUpdate | AnalyticsResult | Periodic analytics update | | shutdown | - | Plugin shutdown |

Triad Deliberation Integration

Integration Points

The Conflict Monitor integrates with the Triad Deliberation Protocol at these points:

1. Proposal Submission - Each proposal is analyzed for conflicts before deliberation
2. During Deliberation - Real-time monitoring of statements for contradictions
3. Voting Phase - Check for conflicts that might block consensus
4. Resolution - Generate suggestions for any detected conflicts

Usage Pattern

// Before deliberation starts
const preCheck = await plugin.monitorTriadDeliberation(deliberation);
if (!preCheck.canProceed) {
  console.log('Blocking conflicts detected:');
  for (const conflict of preCheck.blockingConflicts) {
    const suggestions = plugin.getSuggestions(conflict.id);
    console.log(`- ${conflict.description}`);
    console.log(`  Suggestions: ${suggestions.map(s => s.strategy).join(', ')}`);
  }
}

// During deliberation
plugin.on('conflictDetected', async (conflict) => {
  if (context.isTriadDeliberation) {
    const suggestions = plugin.getSuggestions(conflict.id);
    await notifyTriadMembers(conflict, suggestions);
  }
});

Triad Status

const analytics = plugin.getAnalytics();
console.log(analytics.triadStatus);
// {
//   totalTriadConflicts: 2,
//   byMember: [
//     { member: 'alpha', conflictCount: 1 },
//     { member: 'beta', conflictCount: 0 },
//     { member: 'charlie', conflictCount: 1 }
//   ],
//   blockingDeliberation: false
// }

Configuration

Environment Variables

# Plugin settings
CONFLICT_MONITOR_SENSITIVITY=0.7
CONFLICT_MONITOR_AUTO_DETECT=true
CONFLICT_MONITOR_AUTO_SUGGEST=true
CONFLICT_MONITOR_NOTIFY_CRITICAL=true

# Triad integration
CONFLICT_MONITOR_TRIAD_INTEGRATION=true
CONFLICT_MONITOR_TRIAD_MEMBERS=alpha,beta,charlie

# Analytics
CONFLICT_MONITOR_ANALYTICS=true
CONFLICT_MONITOR_ANALYTICS_INTERVAL=60000

openclaw.json Configuration

{
  "plugins": {
    "conflict-monitor": {
      "enabled": true,
      "path": "./plugins/conflict-monitor",
      "config": {
        "triadIntegration": true,
        "autoDetectConflicts": true,
        "autoGenerateSuggestions": true,
        "notifyOnCritical": true,
        "severityThresholds": {
          "CRITICAL": { "min": 0.85, "max": 1.0 }
        },
        "contextMultipliers": {
          "isTriadDeliberation": 1.3,
          "isEmergency": 1.5
        }
      }
    }
  }
}

Troubleshooting

High false positive rate

1. Reduce sensitivity: sensitivity: 0.5
2. Disable specific detection types
3. Add known contradictions to exclusion list

Missing conflicts

1. Increase sensitivity: sensitivity: 0.8
2. Add custom known contradictions
3. Enable all detection types

Performance issues

1. Reduce maxHistorySize
2. Increase analyticsInterval
3. Disable unused detection types

Testing

# Run tests
npm test

# Run health check
npm run healthcheck

License

MIT

Repository

https://github.com/heretek-ai/heretek-openclaw/tree/main/plugins/conflict-monitor

References

• GAP_ANALYSIS_REPORT.md - Gap analysis
• EXTERNAL_PROJECTS_GAP_ANALYSIS.md - External analysis
• AGENTS.md - Agent documentation
• SKILLS.md - Skills documentation