Protocol Active · v1.1

The Identity Standard
for AI Agents

Verifiable passports, work receipts, and kill switches for autonomous AI workers. Open protocol built on Base L2.

Try Check Your Prompt Read the Spec
Built on Base (Coinbase L2) Open Standard Ed25519 Verified 4 SDKs

How It Works

Three steps to verifiable AI agent identity

Step 1

Register

Agent gets an Ed25519 identity and unique DID, anchored on the blockchain for tamper-proof provenance.

Step 2

Work

Every action produces cryptographically signed receipts, forming an auditable chain of work history.

Step 3

Verify

Anyone can verify an agent's passport and full receipt chain. Trust is transparent and cryptographic.

Live Fleet Overview

Real-time monitoring across the entire agent network

See full dashboard

Works with any agent framework

One command. Universal compatibility.

$ npx aps-mcp-server
Claude GPT MCP LangChain CrewAI

Built-in Primitives

Everything you need for secure autonomous agent operations

Agent Registration

Unique DID + Ed25519 keypair. Verifiable identity anchored on Base L2 blockchain.

Work Receipts

Every action produces a signed receipt. Full audit trail with cryptographic proof of work.

Kill Switch

Soft stop, hard stop, or full revocation. Instant control over any agent in your fleet.

Ready to secure your agents?

Open source. Free forever. No vendor lock-in.

Get Started

Agent Passport Standard v1.1