Five-Layer Model
The Complete HUMΛN Protocol Stack
The HUMΛN protocol is organized into five layers, each providing a distinct set of services to the layer above it. This modular design allows you to use as much or as little of the stack as you need.
Layer Independence
Each layer can be used independently. Need just identity? Use Layer 2 (Passport). Need full orchestration? Use all five layers. Mix and match based on your needs.
Layer 1: Ledger Layer (Provenance)
The foundation of HUMΛN. Every action is recorded in an immutable ledger with cryptographic proofs.
What It Provides
- •Immutable audit trail
- •Provenance chains
- •Cryptographic signatures
- •Query APIs
Use Cases
- •Compliance audits
- •Security investigations
- •Debugging workflows
- •Cost attribution
Layer 2: Identity Layer (Passport)
Cryptographic identity for humans, organizations, and AI agents. Portable, verifiable, and human-owned.
What It Provides
- •Decentralized identifiers (DIDs)
- •Public/private key pairs
- •Delegation tokens
- •Passport management APIs
Use Cases
- •User onboarding
- •Agent authorization
- •Multi-device sync
- •Organization management
Layer 3: Capability Layer (Capability Graph)
A decentralized map of skills and permissions. Agents declare what they can do, humans grant capabilities, and the system routes work accordingly.
What It Provides
- •Capability declarations
- •Grant/verify/revoke APIs
- •Dynamic routing rules
- •Capability discovery
Use Cases
- •Dynamic agent onboarding
- •Multi-skill workflows
- •Access control
- •Agent marketplaces
Layer 4: Orchestration Layer (HumanOS)
The brain of HUMΛN. Coordinates multi-agent workflows, enforces human-in-the-loop approvals, and tracks execution.
What It Provides
- •Workflow orchestration
- •Task routing
- •Human-in-the-loop enforcement
- •Execution plans
Use Cases
- •Multi-agent workflows
- •Approval workflows
- •Escalation handling
- •SLA enforcement
Layer 5: Application Layer (Your Apps)
Everything you build. This is where your agents, workflows, UIs, and integrations live — all built on top of the HUMΛN protocol.
What You Build
- •AI agents & assistants
- •Custom workflows
- •User interfaces
- •Third-party integrations
Examples
- •Invoice processing system
- •Customer support bot
- •Research assistant
- •Data analysis pipeline
How They Work Together
┌──────────────────────────────────────────────────────────────┐ │ CROSS-LAYER FLOW │ ├──────────────────────────────────────────────────────────────┤ │ │ │ USER REQUEST (Layer 5) │ │ "Process 50 invoices" │ │ ↓ │ │ HUMANOS ORCHESTRATION (Layer 4) │ │ - Creates workflow │ │ - Queries Capability Graph for agents │ │ ↓ │ │ CAPABILITY MATCHING (Layer 3) │ │ - Finds agents with invoice_processing capability │ │ - Verifies they're granted and not expired │ │ ↓ │ │ IDENTITY VERIFICATION (Layer 2) │ │ - Checks delegation tokens │ │ - Verifies Passport signatures │ │ ↓ │ │ PROVENANCE LOGGING (Layer 1) │ │ - Logs every step │ │ - Creates provenance chain │ │ - Signs with cryptographic proof │ │ ↓ │ │ AGENT EXECUTION (Layer 5) │ │ - Agent processes invoices │ │ - Returns results to HumanOS │ │ ↓ │ │ RESULT LOGGED (Layer 1) │ │ - Final provenance entry │ │ - Complete audit trail │ │ │ └──────────────────────────────────────────────────────────────┘