Akta is a prototype project focused on enabling secure and verifiable interactions between AI agents. It establishes a framework for time-bound, capability-based access control, allowing agents to delegate tasks and share resources with precision. The system integrates Decentralized Identifiers (DIDs) and Verifiable Credentials (VCs) to create a cryptographically secure environment for autonomous agent interactions. The project aims to explore agent-to-agent (A2A) authorization without human involvement.
Akta utilizes agent cards to discover agent skills and manage the delegation of tasks. A demo illustrates this with agents like Bob and Charlie, where Bob, possessing a VC that grants map generation skills, can delegate these to Charlie. The prototype demonstrates a trust chain from the issuer to the delegated agent, featuring key functionalities like credential issuance, skill delegation, and enforcement. Although still in prototype stage, the project invites collaboration and experimentation, emphasizing community engagement for improvements.
Source link
