Squads
Understanding team-based organization with squads
A Squad is a logical team of AI agents working together. Squads help you organize your agents by purpose, environment, or function, enabling effective collaboration and management.
What is a Squad?
A squad represents:
- An environment (Production, Staging, Development)
- A functional team (DevOps, Security, SRE)
- A project or application team
- Any logical grouping that makes sense for your organization
Why Use Squads?
Organization
Group related agents by purpose:
- All production monitoring agents in one squad
- Each environment or function separate
- Easy to navigate and manage teams
Collaboration
Enable agents to work together:
- Agents in the same squad share context
- Coordinate complex investigations
- Combine specialized capabilities
Isolation
Keep different teams separate:
- Production squad isolated from staging
- Development experiments don't affect production
- Clear boundaries between contexts
Access Control
Manage permissions at the squad level:
- Team members see only their squads
- Different access levels per squad
- Audit trails per team
Flexibility
Organize by what makes sense:
- By environment (Prod, Staging, Dev)
- By function (DevOps, Security, Compliance)
- By application (API Team, Frontend Team)
- By region (US-East Squad, EU-West Squad)
Squad Organization Patterns
By Environment
The most common pattern:
| Squad | Purpose | Agents |
|---|---|---|
Production SRE | Live production monitoring | Web Monitor, DB Guardian, Security Audit |
Staging Team | Pre-production testing | Test Agent, Integration Monitor |
Development Squad | Development support | Dev Helper, Build Monitor |
By Function
Organize by team responsibility:
| Squad | Purpose | Agents |
|---|---|---|
DevOps Team | Infrastructure management | Deployment Agent, Config Monitor |
Security Squad | Security monitoring | Vulnerability Scanner, Audit Agent |
Database Team | Database operations | DB Monitor, Query Optimizer |
By Application
For multi-application organizations:
| Squad | Purpose | Agents |
|---|---|---|
API Squad | API service monitoring | API Monitor, Performance Tracker |
Frontend Team | Frontend monitoring | Web Monitor, Asset Optimizer |
ML Pipeline | ML operations | Model Monitor, Data Validator |
By Region
For geographically distributed operations:
| Squad | Purpose | Agents |
|---|---|---|
US-East Squad | US East operations | Regional Monitor, Failover Agent |
EU-West Squad | European operations | EU Monitor, Compliance Agent |
Combined
Combine patterns as needed:
| Squad | Purpose |
|---|---|
Production API - US-East | API production in US East |
Staging Frontend | Frontend staging environment |
Dev ML Pipeline | ML development team |
Squad Lifecycle
Creating a Squad
- Navigate to Squads in the dashboard
- Click "Create Squad"
- Enter name, environment, and purpose
- Configure optional AI settings
- Save the squad
Squad States
| State | Description |
|---|---|
| Active | Normal operation, agents can be created |
| Paused | No new operations, existing agents preserved |
| Archived | Historical reference, read-only |
Deleting a Squad
Warning: You must delete or move all agents in a squad before deleting the squad itself.
- Remove or reassign all agents
- Go to squad settings
- Click "Delete Squad"
- Confirm deletion
Managing Squad Agents
Adding Agents
- Navigate to your squad
- Click "Create Agent"
- Define agent name and purpose
- Configure system prompt and capabilities
- Link to a deployed node (or link later)
Agent Organization
Within a squad, agents can:
- Work independently on different tasks
- Collaborate on complex investigations
- Share context and findings
- Be reassigned to different nodes as needed
Squad Configuration
Configure squad-level settings:
- Environment: Production, Staging, Development
- Purpose: Team description and goals
- AI Settings: Shared prompts and behavior
- Access Control: Who can manage the squad
Squad Health
Health Indicators
| Status | Meaning |
|---|---|
| Healthy | All agents active and linked |
| Degraded | Some agents unlinked or nodes offline |
| Unhealthy | Most agents unavailable |
| Empty | No agents in squad |
Health Monitoring
The platform tracks:
- Agent link status (linked/unlinked)
- Node connectivity (online/offline)
- Agent activity and responsiveness
- Command execution success rate
Alerting
Configure alerts for:
- Agents becoming unlinked
- Linked nodes going offline
- Squad health status changes
- Failed command executions
Infrastructure Flexibility
Squads are infrastructure-agnostic. Agents can link to nodes running on:
| Infrastructure | Example |
|---|---|
| Virtual Machines | AWS EC2, GCP Compute Engine, Azure VMs |
| Bare Metal | Dedicated servers, on-premise hardware |
| Containers | Docker hosts, Kubernetes nodes |
| Edge Devices | Raspberry Pi, IoT gateways |
| Hybrid | Mix of cloud, on-premise, and edge |
The flexibility of the node-linking architecture allows you to:
- Reassign agents to different servers without recreating them
- Scale infrastructure independently from agent configuration
- Move agents between environments by relinking nodes
Multi-Squad Strategies
Environment Promotion
Organize squads by environment:
Promote changes through squad environments.
Functional Teams
Organize by responsibility:
- DevOps Squad - Infrastructure and deployment
- Security Squad - Security monitoring and compliance
- SRE Squad - Reliability and performance
- Database Squad - Database operations
Regional Operations
Squads for different regions:
- US-East Squad - Primary US operations
- EU-West Squad - European operations
- APAC Squad - Asia-Pacific operations
Each squad has agents linked to regional nodes.
Best Practices
Naming Conventions
Use clear, descriptive squad names:
- Include environment:
Production SRE,Staging DevOps - Include function:
Security Team,Database Squad - Be specific:
API Production Monitoring,Frontend Development
Example: Production API - Security Squad
Documentation
Document your squads:
- Purpose and responsibilities
- Team members with access
- Related systems and services
- Escalation procedures
- Contact information
Agent Management
Within squads:
- Create agents with clear, specific purposes
- Link agents to appropriate nodes
- Monitor agent and node health
- Regularly review and update agent configurations
Limits
Consider subscription limits:
- Number of squads allowed
- Number of agents per squad
- Number of nodes you can deploy
- API rate limits
Cleanup
Regularly review squads:
- Archive unused squads
- Remove obsolete agents
- Unlink agents from decommissioned nodes
- Clean up test and development squads
Next Steps
- Security - Learn about security controls for squads.
- Managing Squads - Manage squads through the dashboard.
- First Deployment - Deploy your first node and create agents.