Settings Location: Navigate to Admin Panel → Settings → Developer → AI Connections (MCP) to manage AI agent configurations and MCP server connections.
MCP Overview
What is Model Context Protocol (MCP)?
AI Agent Framework
Intelligent AutomationStandardized protocol for connecting AI agents with external tools and data sources.
Tool Integration
Extended CapabilitiesEnable AI agents to perform complex tasks by accessing platform APIs and external services.
MCP Benefits
- Enhanced AI Capabilities: AI agents can perform real actions, not just provide advice
- Secure Tool Access: Controlled access to platform features and external services
- Standardized Protocol: Industry-standard protocol for AI-tool integration
- Scalable Architecture: Support for multiple AI agents with different capabilities
- Audit Trail: Complete logging of AI agent actions for transparency and compliance
Agent Types
Available AI Agents
- Builder Agent
- Planner Agent
- Service Agent
Full Administrative AccessComplete CRUD operations and administrative capabilities:
- Community Management: Create, update, and configure communities
- Challenge Creation: Design and publish challenges with actions and rewards
- Reward Management: Create and distribute rewards and recognition
- User Administration: Manage user accounts, roles, and permissions
- Content Management: Create posts, moderate content, and manage feeds
- Analytics Access: Generate reports and analyze platform performance
Agent Configuration
1
Agent Selection
Choose appropriate agent type based on required capabilities and access levels.
2
Tool Access Configuration
Define which tools and APIs each agent can access based on organizational needs.
3
Permission Settings
Configure granular permissions for different types of operations and data access.
4
Security Controls
Set up authentication, rate limiting, and access monitoring for AI agent operations.
MCP Server Configuration
Server Setup
Local MCP Server
On-Premise InstallationDeploy MCP server within your infrastructure for maximum control and security.
Hosted MCP Service
Cloud-Based ServiceUse managed MCP service for simplified setup and maintenance.
Connection Configuration
- Server Connection
- Authentication
- Tool Filtering
MCP Server ConnectivityConfigure connection to MCP server:
- Server URL: MCP server endpoint (typically localhost:4010 for local)
- Authentication: Service authentication keys and tokens
- Connection Security: TLS/SSL configuration and certificate validation
- Health Monitoring: Automatic connection health checks and alerts
- Failover Configuration: Backup server settings for high availability
Tool Management
Available Tools
Community Tools
Community Management
- Create and configure communities
- Manage community settings and branding
- Handle community membership and roles
- Archive and restore communities
Engagement Tools
User Engagement
- Create and manage challenges
- Design and distribute rewards
- Manage achievement systems
- Track user progress and analytics
Content Tools
Content Management
- Create and schedule posts
- Moderate community content
- Manage content templates
- Handle user-generated content
Tool Categories
Administrative Tools
Administrative Tools
System Administration
- User Management: Create, update, delete, and manage user accounts
- Role Management: Assign roles and permissions to users
- Organization Settings: Configure organization-wide settings and preferences
- Team Management: Invite and manage team members
- Audit and Reporting: Generate audit logs and compliance reports
Analytics Tools
Analytics Tools
Data Analysis and Reporting
- User Analytics: Analyze user behavior, engagement, and retention
- Performance Metrics: Track challenge completion rates and user progress
- Revenue Analytics: Monitor reward redemption and points economy
- Custom Reports: Generate tailored reports based on specific criteria
- Data Export: Export analytics data for external analysis
Integration Tools
Integration Tools
External System Integration
- Web Search: Access external information and research capabilities
- Social Media Integration: Connect with social platforms for verification and tracking
- Email Integration: Send notifications and marketing communications
- Webhook Management: Configure and manage webhook integrations
- API Access: Connect with third-party APIs and services
Agent Behavior Configuration
Response Settings
1
Response Style
Configure how AI agents communicate and present information to users.
2
Action Confirmation
Set up approval requirements for AI agent actions that modify data or settings.
3
Error Handling
Define how agents should handle errors and communicate issues to users.
4
Logging Level
Configure detailed logging of agent actions for audit and debugging purposes.
Safety Controls
Safety Mechanisms
AI Safety and ControlMultiple layers of safety controls to prevent unintended actions and ensure responsible AI behavior.
Safety Features
- Action Limits
- Approval Workflows
- Audit Trail
Rate Limiting and RestrictionsControl the frequency and scope of AI agent actions:
- Rate Limiting: Limit number of actions per time period
- Action Quotas: Set daily/monthly limits for different types of operations
- Scope Restrictions: Limit agent actions to specific communities or user segments
- Value Limits: Set maximum values for points, rewards, or other modifications
- Time-Based Restrictions: Limit agent operations to specific time windows
Performance & Monitoring
Agent Performance
Response Analytics
Performance MetricsMonitor AI agent response times, success rates, and user satisfaction scores.
Usage Tracking
Activity MonitoringTrack agent usage patterns, popular tools, and operational efficiency metrics.
System Health
1
Connection Monitoring
Continuous monitoring of MCP server connection health and response times.
2
Error Rate Tracking
Monitor error rates and categorize failures for proactive issue resolution.
3
Resource Usage
Track CPU, memory, and network usage of MCP server and agent operations.
4
Alert Configuration
Set up alerts for performance issues, connection failures, and unusual activity patterns.
Development & Testing
Testing Environment
Agent Testing
Agent Testing
Development and Testing Tools
- Sandbox Environment: Test agent configurations without affecting production data
- Mock Tools: Simulate tool responses for testing agent behavior
- Test Scenarios: Pre-defined test cases for validating agent capabilities
- Performance Testing: Load testing for agent response times and scalability
- Integration Testing: End-to-end testing of agent workflows and tool interactions
Debugging Tools
Agent Debugging
Development SupportComprehensive debugging tools for diagnosing agent behavior and troubleshooting issues.
Security & Compliance
MCP Security
1
Authentication Security
Multi-layered authentication including service keys and user token validation.
2
Network Security
Secure communication protocols and network isolation for MCP connections.
3
Data Protection
Encryption of data in transit and at rest for all MCP communications.
4
Access Auditing
Complete audit trail of all MCP operations and data access for compliance.
Compliance Features
- GDPR Compliance: Privacy-compliant handling of personal data by AI agents
- SOC 2 Compliance: Security controls for AI agent operations and data handling
- Audit Requirements: Complete logging and reporting for regulatory compliance
- Data Minimization: AI agents access only data necessary for requested operations
- Right to Deletion: Support for data deletion requests through AI agent operations
Advanced Configuration
Custom Tool Development
Custom Tools
Specialized CapabilitiesDevelop custom tools for organization-specific requirements and workflows.
Tool SDK
Development FrameworkSoftware development kit for creating and integrating custom MCP tools.
Enterprise Features
- Multi-Tenant Support: Separate MCP configurations for different organizations
- High Availability: Failover and load balancing for enterprise-scale deployments
- Custom Integrations: Specialized integrations with enterprise systems
- Advanced Analytics: Detailed analytics and reporting for agent performance
- 24/7 Support: Enterprise support for MCP configuration and troubleshooting
Getting Started: Begin with the Planner Agent to familiarize yourself with MCP capabilities before enabling full Builder Agent access.
Security First: Always configure appropriate safety controls and approval workflows before enabling AI agents with administrative access.
MCP configuration changes may require MCP server restart. Plan configuration updates during maintenance windows to avoid service disruption.