Trust Triangle
Sekuire is the identity layer for AI agents - the three pillars of the Trust Triangle provide Identity, Visibility, and Control for autonomous software.
The Three Pillars
┌─────────────────┐
│ TRUST TRIANGLE │
└────────┬────────┘
│
┌──────────────┼──────────────┐
│ │ │
┌────────▼───────┐ ┌───▼────────┐ ┌──▼──────────┐
│ IDENTITY │ │ VISIBILITY │ │ CONTROL │
│ │ │ │ │ │
│ Sekuire ID │ │ Beacon │ │ Kill Switch │
│ Ed25519 Keys │ │ Heartbeat │ │ Lease-based │
│ DID Format │ │ Discovery │ │ Revocation │
└────────────────┘ └────────────┘ └─────────────┘Each pillar addresses a fundamental question enterprises have about AI agents:
- Identity - Who is this agent? Is it authentic?
- Visibility - Where is this agent? What is it doing?
- Control - Can I stop this agent if something goes wrong?
Pillar 1: Identity (Sekuire ID)
Every agent gets a deterministic, content-addressable identity derived from its code and configuration. This is cryptographically verifiable and immutable.
- BLAKE3 hash of agent manifest (name, model, system prompt, tools)
- Ed25519 digital signatures for proof-of-authorship
- DID-compatible format for interoperability
- Same agent code always produces the same identity
Pillar 2: Visibility (Beacon)
The Beacon system provides real-time visibility into agent operations through auto-discovery and periodic heartbeats.
- Auto-discovery of running agents across infrastructure
- Periodic heartbeats prove liveness and report status
- Platform detection (Docker, Kubernetes, bare metal, serverless)
- Dashboard integration for fleet-wide visibility
Pillar 3: Control (Kill Switch)
Lease-based liveness with emergency revocation ensures that agents can be stopped at any time, from anywhere.
- Lease-based operation: agents must renew their lease to continue
- Emergency revocation from the dashboard or API
- Workspace-level and agent-level kill controls
- Graceful shutdown with configurable timeout
How the Pillars Connect
The three pillars work together to form a complete trust framework:
- Identity establishes who the agent is before it can operate
- Beacon provides continuous proof that the identified agent is running and healthy
- Kill Switch uses the identity and beacon data to enforce control decisions
# 1. Establish identity
sekuire init --name my-agent
sekuire push
# 2. Enable visibility
sekuire run --heartbeat-interval 30
# 3. Control is automatic (lease-based)
# Dashboard: revoke agent at any timeGovernance Layer
On top of the Trust Triangle, Sekuire provides a governance layer through policies:
- Policy Engine - Define what agents are allowed to do
- Compliance Badges - HIPAA, SOC2, GDPR, PCI-DSS verification
- Audit Logging - Complete trail of agent actions
- Reputation System - Track agent reliability over time
Next Steps
- Sekuire ID - Deep dive into cryptographic identity
- Beacon - Agent discovery and heartbeat
- Kill Switch - Emergency revocation
- Quick Start - Build your first agent