Agents
Create, configure, and manage your AI workforce.
The Agent Management page is where you build and manage your AI workforce. Each agent is a specialised AI worker with its own model, persona, tools, and skills.
Page layout
The Agent Management page has four main tabs:
Your agent directory. Browse all agents as cards, filter by status (Active / Idle), search by name or capability. Each card shows:
Agent name, category, and model
Success rate percentage
Tasks completed and capabilities count
Status indicator (online/offline)
Global agent settings — default models, shared prompts, and system-wide configuration that applies across agents.
Manage how agents work together — team assignments, delegation rules, and inter-agent communication settings.
Multi-step workflows that chain agents together. Create, edit, and manage recipes that coordinate multiple agents to complete complex tasks.
Stats bar
At the top of the page, four metric cards give you an at-a-glance view:
Total Agents
Number of agents in your workspace
Active Agents
Agents currently online and processing
Categories
How many distinct agent types you have
Avg Performance
Aggregate success rate across all agents
Quick start: Create your first agent
Click + Create Agent in the top right
Choose a template (Code Reviewer, QA Engineer, etc.) or start blank
Set a name, select a model, and assign a persona
Add tools and skills
Click Create
Your agent appears on the roster immediately and is ready to receive messages.
Start with a template. You can always customise the agent later from the Agent Details panel.
Subpages
Browsing, filtering, and managing your agents
Step-by-step agent creation guide
The per-agent detail panel (Overview, Performance, Plugins, Reports)
Global agent configuration
Multi-agent teamwork settings
Creating and managing multi-step workflows
Last updated