Skip to content

Conformance Profiles

Choose Your Profile

ARAL defines three conformance profiles that progressively build on each other, from foundational to fully integrated.

📦

ARAL-CORE

Layers 1–5

The foundation for standalone autonomous agents with all essential capabilities.

68 Requirements

🎭

ARAL-ORCH

Layers 1–6 (CORE + L6)

Extends CORE with multi-agent orchestration for team coordination.

94 Requirements

🌐

ARAL-INTEROP

Layers 1–7 (Complete)

Full interoperability with standardized protocol for cross-system integration.

120 Requirements

FeatureARAL-COREARAL-ORCHARAL-INTEROP
Runtime (L1)
Memory (L2)
Capabilities (L3)
Reasoning (L4)
Persona (L5)
Orchestration (L6)✅ NEW
Protocol (L7)✅ NEW
Total Requirements6894120
ComplexityLowMediumHigh
Time to Implement1-2 weeks3-4 weeks6-8 weeks

The foundation profile for building standalone autonomous agents. Implements layers 1 through 5, providing all essential capabilities for an intelligent agent to operate independently.

💬

Chatbots & Virtual Assistants

Single-user conversational agents with memory and personality.

Customer service bot, personal productivity assistant, FAQ bot

🤖

Task Automation

Agents that automate specific workflows or processes.

Email classifier, data processor, report generator

📊

Analytics Agents

Agents that analyze data and provide insights.

Sales analytics, log analyzer, trend detector

🎯

Single-Purpose Tools

Focused agents with specific expertise.

Code reviewer, content moderator, translator

🤖

Autonomous Operation

Operates independently without external coordination

💾

Stateful Interactions

Maintains context across conversations

Action Execution

Can perform actions through capabilities

🧠

Reasoning Engine

Makes decisions based on context

👤

Personality

Consistent behavior and tone

📊

Resource Management

Efficient compute and memory use

✅ Implementation Checklist

  • ✓ Implement Runtime Layer (L1) - Platform abstraction, resource management
  • ✓ Implement Memory Layer (L2) - Context window, persistence, state
  • ✓ Implement Capabilities Layer (L3) - Action registry, validation, execution
  • ✓ Implement Reasoning Layer (L4) - Decision engine, planning
  • ✓ Implement Persona Layer (L5) - Identity, traits, constraints
  • ✓ Run: npx @aral-standard/validator —profile ARAL-CORE
{
"aral": {
"version": "1.0",
"profile": "ARAL-CORE",
"conformance": "strict"
},
"agent": {
"id": "customer-support-bot",
"name": "Customer Service Assistant",
"version": "1.0.0"
},
"layers": {
"runtime": { /* L1 config */ },
"memory": { /* L2 config */ },
"capabilities": { /* L3 config */ },
"reasoning": { /* L4 config */ },
"persona": { /* L5 config */ }
}
}

Extends ARAL-CORE with orchestration capabilities (Layer 6) for coordinating multiple agents into teams and workflows.

┌─────────────────────────────────────────┐
│ ✓ L6 Orchestration Multi-Agent Coord │
├─────────────────────────────────────────┤
│ ✓ L5 Persona Identity & Contract│
├─────────────────────────────────────────┤
│ ✓ L4 Reasoning Logic & Decisions │
├─────────────────────────────────────────┤
│ ✓ L3 Capabilities Available Actions │
├─────────────────────────────────────────┤
│ ✓ L2 Memory State & Context │
├─────────────────────────────────────────┤
│ ✓ L1 Runtime Execution Platform │
└─────────────────────────────────────────┘
users 👥 Multi-Agent Teams

Multiple specialized agents working together.

Example: Software development team (architect, coder, tester), research team, content creation team

diagram-project 🔄 Complex Workflows

Sequential or parallel agent execution patterns.

Example: Document processing pipeline, approval workflows, multi-stage analysis

hexagon 🐝 Agent Swarms

Large numbers of agents coordinated towards a common goal.

Example: Web scraping swarm, distributed testing, market monitoring

sitemap 🎯 Hierarchical Systems

Supervisor agents managing worker agents.

Example: Project manager agent coordinating specialists, quality control systems

All ARAL-CORE requirements (68) +
26 additional orchestration requirements = 94 total

Implement ARAL-CORE

Complete all L1-L5 requirements first

Implement Orchestration Layer (L6)

Team management, workflow engine, coordination protocols

Define Team Structures

Roles, responsibilities, communication patterns

Implement Workflow DSL

Language for defining agent workflows

Add Monitoring

Track team performance and agent health

Validate Conformance

Run npx @aral-standard/validator --profile ARAL-ORCH

{
"aral": {
"version": "1.0",
"profile": "ARAL-ORCH",
"conformance": "strict"
},
"agent": {
"id": "dev-team-orchestrator",
"name": "Software Development Team",
"version": "1.0.0"
},
"layers": {
"runtime": { /* L1 config */ },
"memory": { /* L2 config */ },
"capabilities": { /* L3 config */ },
"reasoning": { /* L4 config */ },
"persona": { /* L5 config */ },
"orchestration": {
"team": {
"members": [
{"id": "architect", "role": "designer"},
{"id": "developer", "role": "implementer"},
{"id": "tester", "role": "validator"}
]
},
"workflows": [
{
"id": "feature-development",
"steps": [
{"agent": "architect", "action": "design"},
{"agent": "developer", "action": "implement"},
{"agent": "tester", "action": "validate"}
]
}
]
}
}
}

The complete profile implementing all 7 layers, including standardized interoperability protocols (Layer 7) for cross-system communication.

┌─────────────────────────────────────────┐
│ ✓ L7 Protocol Interoperability │
├─────────────────────────────────────────┤
│ ✓ L6 Orchestration Multi-Agent Coord │
├─────────────────────────────────────────┤
│ ✓ L5 Persona Identity & Contract│
├─────────────────────────────────────────┤
│ ✓ L4 Reasoning Logic & Decisions │
├─────────────────────────────────────────┤
│ ✓ L3 Capabilities Available Actions │
├─────────────────────────────────────────┤
│ ✓ L2 Memory State & Context │
├─────────────────────────────────────────┤
│ ✓ L1 Runtime Execution Platform │
└─────────────────────────────────────────┘
building 🏢 Enterprise Ecosystems

Large-scale agent systems across departments and vendors.

Example: Enterprise-wide AI platform, cross-departmental automation

handshake 🌐 Cross-Vendor Integration

Agents from different vendors working together.

Example: Microsoft agents + LangChain agents + custom agents

store 🛒 Agent Marketplaces

Discovering and invoking third-party agents.

Example: AI agent marketplace, capability exchange

network-wired 🔗 Federated Networks

Distributed agent networks across organizations.

Example: Supply chain coordination, inter-company workflows

All ARAL-ORCH requirements (94) +
26 protocol requirements = 120 total

Implement ARAL-ORCH

Complete all L1-L6 requirements first

Implement Protocol Layer (L7)

Message formats, discovery, handshake, error handling

Add Transport Support

HTTP, WebSocket, gRPC, or custom transports

Implement Security

mTLS, JWT, RBAC, encryption at rest and in transit

Add Versioning

Handle protocol version negotiation

Validate Conformance

Run npx @aral-standard/validator --profile ARAL-INTEROP

{
"aral": {
"version": "1.0",
"profile": "ARAL-INTEROP",
"conformance": "strict"
},
"agent": {
"id": "enterprise-agent-hub",
"name": "Enterprise Agent Hub",
"version": "1.0.0"
},
"layers": {
"runtime": { /* L1 config */ },
"memory": { /* L2 config */ },
"capabilities": { /* L3 config */ },
"reasoning": { /* L4 config */ },
"persona": { /* L5 config */ },
"orchestration": { /* L6 config */ },
"protocol": {
"version": "1.0",
"transports": ["https", "wss"],
"discovery": {
"enabled": true,
"registry": "https://agents.example.com/registry"
},
"security": {
"authentication": "mtls",
"encryption": "tls1.3"
}
}
}
}

Use this decision tree to select the right profile:

No

Yes

No

Yes

Do you need multi-agent coordination?

ARAL-CORE

Do you need cross-system interoperability?

ARAL-ORCH

ARAL-INTEROP

✓ Standalone agents

✓ Single purpose

✓ Fastest implementation

✓ Agent teams

✓ Workflows

✓ Internal systems

✓ Cross-vendor

✓ Enterprise scale

✓ Marketplaces

Note

Recommendation: Start with ARAL-CORE and upgrade to higher profiles as your requirements evolve. Each profile is a proper superset of the previous one.

Start with ARAL-CORE

Build a solid foundation with L1-L5

⏱️ Time: 1-2 weeks
📊 Complexity: Low

Upgrade to ARAL-ORCH

Add Layer 6 when you need coordination

⏱️ Time: +2-3 weeks
📊 Complexity: Medium

Upgrade to ARAL-INTEROP

Add Layer 7 for full interoperability

⏱️ Time: +3-4 weeks
📊 Complexity: High

Each profile has specific validation requirements:

Terminal window
# Validate CORE conformance
npx @aral-standard/validator \
--profile ARAL-CORE \
--manifest ./agent.json \
--output report.html

Tests:

  • ✅ All L1-L5 requirements
  • ✅ Schema validation
  • ✅ Security checks
  • ✅ Performance benchmarks
Terminal window
# Validate ORCH conformance
npx @aral-standard/validator \
--profile ARAL-ORCH \
--manifest ./agent.json \
--test-workflows \
--output report.html

Tests:

  • ✅ All CORE tests
  • ✅ L6 orchestration requirements
  • ✅ Team coordination
  • ✅ Workflow execution
Terminal window
# Validate INTEROP conformance
npx @aral-standard/validator \
--profile ARAL-INTEROP \
--manifest ./agent.json \
--test-protocol \
--output report.html

Tests:

  • ✅ All ORCH tests
  • ✅ L7 protocol requirements
  • ✅ Cross-agent communication
  • ✅ Security protocols