Preview Release: Early-access preview, information and pricing subject to change.

Nine Agent Types

Classify agents by their primary function to determine appropriate governance requirements, oversight levels, and control frameworks.

Assistant Agent

Assists humans with content creation, analysis, research, and knowledge work

Typical Use Cases

  • • Content writing
  • • Research synthesis
  • • Data analysis
  • • Code assistance

Typical Governance Tier

Tier 1: Limited

Typical Controls

  • • Human-in-loop
  • • Read-only access
  • • Public data only
  • • Session memory

Workflow Agent

Executes predefined business processes and workflows with clear steps

Typical Use Cases

  • • Invoice processing
  • • Customer onboarding
  • • Data migration
  • • Report generation

Typical Governance Tier

Tier 2: Controlled

Typical Controls

  • • Defined workflows
  • • Limited execution authority
  • • Approval gates
  • • Audit logging

Routing Agent

Determines where work, requests, or data should be sent next

Typical Use Cases

  • • Customer support routing
  • • Task assignment
  • • Load balancing
  • • Escalation logic

Typical Governance Tier

Tier 1: Limited

Typical Controls

  • • Recommendation only
  • • Human review
  • • Routing rules
  • • Stateless

Monitoring Agent

Observes systems, events, and data to detect changes or anomalies

Typical Use Cases

  • • System health monitoring
  • • Security alerts
  • • Compliance scanning
  • • Event detection

Typical Governance Tier

Tier 2: Controlled

Typical Controls

  • • Read-only
  • • Alert thresholds
  • • Human-on-loop
  • • No execution authority

Decision Support Agent

Provides recommendations and insights to humans for decision making

Typical Use Cases

  • • Risk assessment
  • • Investment analysis
  • • Diagnostic support
  • • Strategic planning

Typical Governance Tier

Tier 2: Controlled

Typical Controls

  • • Recommendation only
  • • Source citation
  • • Confidence scoring
  • • Human final decision

Tool-Using Agent

Selects and uses approved tools, APIs, or functions to accomplish tasks

Typical Use Cases

  • • API orchestration
  • • Multi-tool workflows
  • • System integration
  • • Automated testing

Typical Governance Tier

Tier 3: Material

Typical Controls

  • • Tool allowlist
  • • Parameter validation
  • • Rate limiting
  • • Human-on-loop

Execution Agent

Executes actions with real-world consequences (financial, legal, operational)

Typical Use Cases

  • • Trade execution
  • • Payment processing
  • • Resource provisioning
  • • Contract signing

Typical Governance Tier

Tier 3: Material

Typical Controls

  • • Pre-approval required
  • • Transaction limits
  • • Audit trail
  • • Rollback capability

Orchestration Agent

Coordinates multiple agents, systems, and workflows across boundaries

Typical Use Cases

  • • Multi-agent coordination
  • • System-of-systems management
  • • Complex workflow execution
  • • Resource allocation

Typical Governance Tier

Tier 3: Material

Typical Controls

  • • Delegation tracking
  • • Cascading approvals
  • • Failure handling
  • • Human oversight

Strategic Agent

Pursues broad objectives with emergent, adaptive, or creative paths to goals

Typical Use Cases

  • • Business strategy
  • • Research exploration
  • • Competitive analysis
  • • Innovation projects

Typical Governance Tier

Tier 4: Critical

Typical Controls

  • • Executive approval
  • • Continuous monitoring
  • • Strict boundaries
  • • Frequent review

Ready to Classify Your Agents?

Start registering and governing your AI agents with the AGEI framework