Description
Handshake.AI is an open protocol designed to provide cryptographic identity, verifiable delegation, and signed action receipts for AI agents. This protocol ensures that AI agents and the systems they interact with can prove who performed an action, under whose authority, and what exactly happened. By using cryptographic identity, every agent is assigned a keypair and a verifiable credential, allowing any party to verify the agent without needing to contact the original deployer. The protocol also supports a provable chain of consent, where each step in a user-agent interaction carries a signed delegation token, ensuring that the entire chain of actions can be traced back to a human principal.
Read More
Handshake.AI's protocol is designed to be invisible to the agent author, with an SDK that wraps existing agent frameworks. This SDK adds the handshake to outbound calls and produces signed receipts as a side effect of normal operation. The protocol is compatible with existing systems, wrapping an attestation envelope around the protocols that define how agents communicate with tools and each other. This means that the existing stack remains unchanged, and the audit layer is added on top.
The protocol is open-source and MIT-licensed, with a hosted Registry and enterprise Console available for deployment. Handshake.AI is currently working with design partners to refine the protocol and infrastructure. The company is also building towards a compliance roadmap that includes SOC2 Type I and II, ISO 27001, and FedRAMP Moderate certifications. Handshake.AI is particularly relevant for enterprises that require a robust audit and assurance posture for their AI systems, especially in regulated environments.
The protocol is designed to address several key challenges in AI agent deployment, including unverifiable agent identity, unsafe delegation of authority, and the lack of a signed audit trail. By providing a comprehensive solution to these issues, Handshake.AI aims to facilitate the secure and compliant deployment of AI agents across various industries.
- CategorySecurity & Privacy
- Pricing ModelOpen Source
- DeploymentCloud, Self-hosted
- PlatformsCloud
Handshake.AI's Core Features
Cryptographic agent identity
Verifiable delegation
Signed action receipts
Open protocol
Provable chain of consent
Negotiated, scoped capabilities
Signed action attestations
SDK for agent frameworks
Compatibility with existing protocols
Hosted Registry and enterprise Console
How to use Handshake.AI?
Install SDK: Integrate the Handshake SDK into your agent framework
Configure Protocol: Set up cryptographic identities and delegation tokens
Initiate Handshake: Agents perform a handshake before meaningful actions
Verify Actions: Use signed receipts to verify actions and authority
Handshake.AI's Use Cases
- Agent Identity Verification
- Delegation Management
- Audit Trail Creation
- Compliance Assurance
- Secure Agent Deployment
