Agent Execution Guide
Monitor and manage AI development agents in your Lovelace workspace
What are Agents?
Agents in Lovelace are computational units that execute specific development tasks—code analysis, testing, documentation generation, and automation workflows. Unlike conversational AI assistants, agents are:
- Task-focused - Designed to execute specific development work
- Stateless - Run independently without persistent context
- Managed by a daemon - Controlled through the localhost daemon
- Lightweight - Can be distributed and scaled
Agents are the execution layer of the Lovelace platform. When you use the CLI or interact with the platform, agents are the infrastructure performing the actual work.
Prerequisites
Before using agent commands, ensure the localhost daemon is running. This daemon manages the agent lifecycle and execution environment.
If you haven't set up the localhost daemon yet, follow the installation guide to get started.
Listing Running Agents
View all agents currently running in your workspace:
# List all running agents
lovelace agents list
# Get JSON output for scripting
lovelace agents list --json
Example Output
📋 Running Agents (3)
1. code-analysis-agent
Task: static-analysis
Status: idle
2. test-generation-agent
Task: unit-test-generation
Status: processing
3. documentation-agent
Task: api-docs
Status: idle
The list shows:
- Agent ID - Unique identifier for the agent
- Task - Current or assigned task
- Status - Current state (idle, processing, error)
Use --json output when you need to parse the results programmatically or integrate with other tools.
Checking System Status
Agent System Status
Get an overview of your agent infrastructure:
lovelace agents status
This shows:
- Whether agents are running
- Total number of active agents
- Overall system health
- Agent environment state
Daemon Status
Check the health and metrics of the daemon that manages agents:
# Basic daemon status
lovelace agents daemon status
# Detailed metrics and information
lovelace agents daemon status --verbose
# Include health check results
lovelace agents daemon status --health
# Enable debug logging
lovelace agents daemon status --debug
Understanding Daemon Output
The daemon status output includes:
- Process ID (PID) - The daemon's system process ID
- Uptime - How long the daemon has been running
- Agent Environment - Whether the runtime environment is active
- Health Status - System health and performance metrics
Workflow: Monitor Your Agents
Here's a typical workflow for monitoring agents:
-
Start with system status - Verify the daemon is running:
bashlovelace agents status -
Check running agents - See which agents are active:
bashlovelace agents list -
Monitor daemon health - For production systems, check daemon metrics:
bashlovelace agents daemon status --verbose
Coming Soon
The following features are currently in development:
Agent Logs
lovelace agents logs <agent-id>
View detailed logs from a specific agent to debug issues or monitor progress.
Agent Information
lovelace agents info <agent-id>
Get detailed information about a specific agent, including configuration, capabilities, and recent activity.
These commands will be available once the localhost daemon supports log streaming and detailed agent inspection.
Troubleshooting
No Agents Listed
Problem: lovelace agents list returns "No agents are currently running"
Cause: Daemon is running but no agents have been started
Solution:
- Check daemon status:
lovelace agents daemon status - Verify the daemon is active (should show running status)
- Agents start automatically when tasks are assigned through the platform or CLI chat
Daemon Not Running
Problem: Connection errors when running agent commands
Cause: The localhost daemon is not running
Solution:
# Check daemon status
lovelace agents daemon status
# If not running, start the daemon through localhost CLI
# Consult your installation guide for daemon startup instructions
Connection Issues
Problem: "Failed to connect to agent system" errors
Cause: Daemon is running but not responding
Solution:
- Check daemon health:
lovelace agents daemon status --health - Review daemon logs for errors
- Restart the daemon
- Verify network connectivity on your system
Next Steps
- Interactive Chat - Use agents in conversational mode
- MCP Integration - Connect agents with external tools
- Workspace Management - Manage your workspace