Examples

Real-world agent manifests showing progressive profile adoption

Each example builds upon the previous, demonstrating how to progressively adopt JSON Agents profiles based on your needs — from simple core agents to complex multi-agent orchestrations.

Core Profile Only

Minimal agent manifest with identity, capabilities, and tools

core
View JSON

What's Included:

  • Agent identity (name, version, description)
  • Capabilities (summarization, routing, etc.)
  • Tool definitions
  • Context configuration

Use case: Simple standalone agents without runtime or governance requirements

Core + Exec Profile

Agent with runtime and execution environment configuration

coreexec
View JSON

Adds to Core:

  • Runtime platform specification
  • Model configuration (provider, version, parameters)
  • Environment variables
  • Resource requirements

Use case: Production-ready agents with specific deployment requirements

Core + Exec + Gov Profile

Agent with security policies, observability, and governance

coreexecgov
View JSON

Adds to Core + Exec:

  • Security policies and access control
  • Sandbox configuration
  • Audit logging
  • Observability (metrics, traces, logs)
  • Cryptographic signatures

Use case: Enterprise agents requiring compliance, security, and monitoring

Full Profile Stack

Complete multi-agent system with orchestration topology

coreexecgovgraph
View JSON

Adds to Core + Exec + Gov:

  • Multi-agent topology definition
  • Node and edge specifications
  • Conditional routing logic
  • Entry and exit points
  • Inter-agent communication flows

Use case: Complex agent networks, workflows, and collaborative systems

Related Resources

Continue exploring JSON AGENTS