Dashboard Guide

The AgentHub dashboard is your central control panel for managing agents, monitoring instances, tracking analytics, and configuring your organization. This guide covers all dashboard features and workflows.

Dashboard Overview

The dashboard is organized into several main sections:

• Agents - Manage your agent blueprints and templates
• Instances - Monitor running agent deployments
• Analytics - Track performance, costs, and usage metrics
• Organization - Team management and billing
• Settings - API keys, preferences, and configuration

Getting Started

Signing In

1. Navigate to AgentHub Dashboard
2. Click "Sign in with GitHub"
3. Authorize AgentHub to access your GitHub account
4. Complete organization setup if this is your first visit

First-Time Setup

Organization Creation:

• Choose an organization name (can be changed later)
• Invite team members via email
• Configure billing settings

GitHub Integration:

• Grant repository access for the repos you want to use
• Configure webhook permissions for automatic builds
• Set up SSH keys if using private repositories

Agents Section

Agents Overview

The Agents section shows all your published agent blueprints:

Key Information Displayed:

• Agent name and description
• GitHub repository and branch
• Number of builds and active instances
• Last deployment date
• Status (active, archived, build failed)

Quick Actions:

• Deploy - Create new instance from latest build
• Build - Trigger new container build
• Settings - Configure agent parameters
• Analytics - View agent-specific metrics

Creating Agents

From GitHub Repository

1. Click "New Agent" or "Deploy Your Agent"
2. Select Repository:
   • Choose from your authorized GitHub repos
   • Use search to find specific repositories
   • Select branch (defaults to main)
3. Configure Agent:
   • Set agent name (displayed to users)
   • Add description/summary
   • Configure environment variables
   • Set resource limits (CPU, memory)
4. Deployment Settings:
   • Choose deployment type (HTTP, long-running, stdin)
   • Set port and health check endpoint
   • Configure auto-scaling options
5. Review and Deploy:
   • Preview configuration
   • Click "Create Agent" to start initial build

Auto-Configuration

AgentHub automatically detects:

• Framework (Python, Node.js, Go, Rust, etc.)
• Dependencies (requirements.txt, package.json, etc.)
• Entry point (main.py, index.js, etc.)
• Port (from code analysis or common defaults)
• Container configuration (optimized Dockerfile generation)

Managing Agents

Agent Settings

Basic Configuration:

• Name and description
• GitHub repository and default branch
• Public visibility settings

Runtime Configuration:

• Port and health check path
• Working directory and startup command
• Long-running vs. cold-start behavior
• HTTP endpoint enablement

Environment Variables:

TARGET_QUERY=renewable energy stocks
GEMINI_API_KEY=(leave empty for security)
LOG_LEVEL=INFO
CACHE_TTL=3600

Resource Limits:

• CPU allocation (0.1 - 4.0 cores)
• Memory allocation (128Mi - 8Gi)
• Burst limits for peak performance

Agent Actions

Build Management:

• Trigger manual builds
• View build history and logs
• Manage build artifacts and versions

Instance Deployment:

• Create test instances
• Deploy to production
• Configure auto-scaling rules

Sharing and Distribution:

• Generate shareable URLs
• Create custom landing pages
• Configure public visibility

Agent Analytics

Build Metrics:

• Build success/failure rates
• Build time trends
• Image size optimization
• Error analysis

Usage Metrics:

• Total instances created
• Active instance count
• Geographic distribution
• User engagement

Performance Metrics:

• Average response times
• Resource utilization
• Cost per instance
• Uptime statistics

Instances Section

Instances Overview

Monitor all running agent deployments across your organization:

Instance List View:

• Instance name and status indicator
• Parent agent and build version
• Resource usage (CPU, memory)
• Uptime and restart count
• Public URL and access links

Status Indicators:

• 🟢 Running - Instance is healthy and serving requests
• 🟡 Starting - Instance is being provisioned
• 🔴 Error - Instance has failed or crashed
• ⚪ Stopped - Instance has been terminated

Instance Management

Starting New Instances

1. From Agent Page:
   • Go to Agents → Select agent → Click "Start Instance"
   • Choose build version (latest or specific version)
   • Configure instance-specific settings
2. From Instances Page:
   • Click "New Instance"
   • Select agent blueprint
   • Configure deployment parameters

Instance Configuration:

{
  "name": "production-trading-bot",
  "resources": {
    "cpu": "1.0",
    "memory": "2Gi",
    "replicas": 2
  },
  "env_vars": {
    "API_KEY": "your-production-key",
    "TRADING_MODE": "live",
    "RISK_LEVEL": "moderate"
  }
}

Monitoring Instances

Real-time Logs:

• Stream live application logs
• Filter by log level (DEBUG, INFO, WARN, ERROR)
• Search and export log data
• Set up log alerts

Performance Metrics:

• CPU and memory usage graphs
• Request rate and response time
• Network I/O and disk usage
• Custom application metrics

Health Status:

• Endpoint health checks
• Container restart events
• Resource limit violations
• Error rate monitoring

Instance Actions

Runtime Control:

• Start/Stop - Control instance lifecycle
• Restart - Restart containers (preserves data)
• Scale - Adjust replica count
• Update - Deploy new configuration

Configuration Management:

• Update environment variables
• Modify resource allocations
• Change scaling parameters
• Configure health checks

Troubleshooting:

• Access container shell
• Download log files
• View resource usage history
• Generate debug reports

Instance Logs and Debugging

Log Viewer

Features:

• Real-time log streaming
• Historical log search
• Multi-level filtering
• Export and download options

Log Levels:

[DEBUG] 2024-01-15 10:30:00 - Processing market data for AAPL
[INFO]  2024-01-15 10:30:05 - Generated 5 trading signals
[WARN]  2024-01-15 10:30:10 - API rate limit approaching
[ERROR] 2024-01-15 10:30:15 - Failed to fetch news data: timeout

Search and Filtering:

• Text search across all logs
• Filter by time range
• Filter by log level
• Regular expression support

Interactive Debugging

Container Shell Access:

# Access running container
/app$ ps aux
/app$ ls -la
/app$ python -c "import sys; print(sys.path)"

Command Execution:

• Run diagnostic commands
• Inspect file system
• Check process status
• Debug environment issues

Analytics Section

Performance Analytics

Response Time Metrics:

• Average, median, 95th percentile response times
• Response time distribution graphs
• Performance trends over time
• Geographic latency analysis

Throughput Metrics:

• Requests per minute/hour/day
• Peak traffic identification
• Load pattern analysis
• Capacity planning insights

Error Analysis:

• Error rate trends
• Error categorization (4xx, 5xx, timeouts)
• Most common error messages
• Error correlation with deployments

Cost Analytics

Resource Costs:

• Compute costs (CPU, memory, storage)
• Network transfer costs
• Container registry costs
• Total cost breakdown

LLM API Costs:

• Per-provider cost tracking (OpenAI, Anthropic, etc.)
• Token usage and billing
• Cost optimization recommendations
• Budget alerts and limits

Cost Attribution:

• Per-agent cost analysis
• Per-instance cost tracking
• Team/user cost allocation
• Billing period summaries

Usage Analytics

Instance Metrics:

• Total instances created
• Active vs. stopped instances
• Instance lifetime statistics
• Resource utilization efficiency

User Engagement:

• Unique users per agent
• User retention rates
• Feature adoption metrics
• Geographic user distribution

Build Analytics:

• Build frequency and success rates
• Build time optimization
• Image size trends
• Deployment frequency

Custom Dashboards

Dashboard Creation:

• Drag-and-drop dashboard builder
• Custom metric visualization
• Real-time data updates
• Shared team dashboards

Available Widgets:

• Time series graphs
• Status indicators
• Alert panels
• Cost summaries
• Performance scorecards

Organization Management

Team Management

Member Roles:

• Owner - Full administrative access
• Admin - Manage agents and instances
• Developer - Deploy and modify agents
• Viewer - Read-only access

Inviting Members:

1. Go to Settings → Organization → Members
2. Click "Invite Member"
3. Enter email address and select role
4. Send invitation
5. Member receives email with setup link

Permission Management:

• Role-based access control
• Resource-level permissions
• API key scope limitations
• Audit log tracking

Billing and Usage

Subscription Management:

• View current plan and usage
• Upgrade/downgrade plans
• Manage payment methods
• Download invoices

Usage Monitoring:

• Real-time usage tracking
• Historical usage reports
• Budget alerts and limits
• Cost optimization recommendations

Resource Quotas:

• Agent limits per organization
• Instance limits per user
• Build quota management
• Storage limit monitoring

Settings and Configuration

API Keys

Creating API Keys:

1. Go to Settings → API Keys
2. Click "Generate New Key"
3. Set name and expiration
4. Choose scope (organization vs. user level)
5. Copy key (shown only once)

Key Management:

• View active keys and last used dates
• Set expiration dates
• Revoke compromised keys
• Monitor key usage

GitHub Integration

Repository Access:

• Manage repository permissions
• Add new repositories
• Configure webhook endpoints
• Set up SSH key authentication

Build Automation:

• Auto-build on push events
• Branch-specific build rules
• PR build previews
• Tag-based releases

Preferences

UI Customization:

• Dark/light theme
• Dashboard layout preferences
• Notification settings
• Time zone configuration

Default Settings:

• Default resource allocations
• Environment variable templates
• Build configuration defaults
• Instance naming conventions

Mobile Dashboard

Mobile App Features

Core Functionality:

• Monitor instance status
• View real-time logs
• Start/stop instances
• Receive push notifications

Quick Actions:

• Emergency instance stop
• Scale up/down instances
• View critical alerts
• Access support chat

Mobile Optimization

Responsive Design:

• Touch-friendly interface
• Swipe gestures for actions
• Optimized for small screens
• Offline status indicators

Keyboard Shortcuts

Navigation

• Ctrl/Cmd + K - Quick search
• G + A - Go to Agents
• G + I - Go to Instances
• G + D - Go to Dashboard
• G + S - Go to Settings

Actions

• N - Create new agent/instance
• R - Refresh current view
• ? - Show keyboard shortcuts
• Esc - Close modals/panels

Troubleshooting Common Issues

Agent Deployment Problems

Build Failures:

1. Check build logs for specific errors
2. Verify GitHub repository access
3. Ensure dependencies are properly specified
4. Check Dockerfile syntax if using custom builds

Configuration Issues:

1. Verify environment variables are set correctly
2. Check resource limits are sufficient
3. Ensure port configuration matches your application
4. Validate health check endpoint

Instance Issues

Startup Failures:

1. Check instance logs for error messages
2. Verify environment variables contain correct values
3. Ensure sufficient resources are allocated
4. Check build compatibility with runtime

Performance Problems:

1. Monitor resource usage graphs
2. Check for memory leaks or CPU spikes
3. Review log patterns for errors
4. Consider scaling up resources or replicas

Dashboard Access Problems

Login Issues:

1. Clear browser cache and cookies
2. Verify GitHub account has correct permissions
3. Check if organization membership is active
4. Try incognito/private browsing mode

Performance Issues:

1. Check internet connection stability
2. Try refreshing the page
3. Clear browser cache
4. Contact support if problems persist

---

Next: Analytics Platform Guide →