Agent Name Service A Linux Foundation project
Verifiable identityfor AI agents.
ANS is an open specification for naming, discovering, and authenticating AI agents. It connects domain names, certificates, and transparency records so peers can verify who they are communicating with.
The published specifications are drafts under active development.
图表:Domain Transparency log Peer
payments.example.com
ans://v1.0.0.payments.example.com
Watch the overview
What is ANS?
A short introduction to agent identity and why it matters when AI agents interact.
图表:1 min 52 sec · English captions available Download video
Download the ANS overviewto watch it in your video player.
Implementations
Build with ANS.
Run the Registration Authority and Transparency Log locally, inspect a registration, and verify its receipt. Use an SDK to bring verification into your application.
Reference implementation
A runnable registration lifecycle, transparency log, and offline verifier.
Go
Registry and transparency clients, verification, and the ANS CLI.
Rust
API clients, shared types, and verification libraries with rustls support.
Java
Registration, discovery, cryptography, and agent client modules.
Browse implementations, integrations, and their supported scope →
Community & governance
Participate in the project.
ANS develops through public specifications, code, and decisions.
Technical Steering Committee
The TSC guides project-wide technical direction, governance, and escalated decisions.
Maintainers & reviewers
Repository teams review changes, maintain implementations, and take responsibility for releases.
A place to get involved
Contribute code, improve documentation, test an integration, or help review a proposal.
ANS establishes identity. Applications still decide what a peer may do.