ans-spec

mcp
Guvenlik Denetimi
Gecti
Health Gecti
  • License — License: NOASSERTION
  • Description — Repository has a description
  • Active repo — Last push 0 days ago
  • Community trust — 33 GitHub stars
Code Gecti
  • Code scan — Scanned 7 files during light audit, no dangerous patterns found
Permissions Gecti
  • Permissions — No dangerous permissions requested

Bu listing icin henuz AI raporu yok.

SUMMARY

An Agent Name Service (ANS) registry, based on https://datatracker.ietf.org/doc/html/draft-narajala-ans-00

README.md

Agent Name Service (ANS)

My company's payment agent receives an instruction to wire $50,000 to a supplier's
invoicing agent. The payment agent must decide: does this agent actually belong to
the supplier, or has someone stood up a convincing fake? And if the site is
legitimate, has the supplier's code changed since my last transaction?

ANS answers both questions. It anchors every agent's identity to a domain name
whose ownership has been verified, issues a version-bound certificate that records
which code is running, and seals every lifecycle event into a Transparency Log
where entries cannot be altered after the fact. The standard is open and the
services are federated, but the identity always anchors to a domain name.

Specifications

The ANS protocol is split across seven layered specifications under spec/:

Layer File Topic
ANS-0 spec/ans-0-identity-anchor.md The proof-of-control gate, Verified Identities, identity profiles
ANS-1 spec/ans-1-registration.md Registration aggregate, lifecycle, event set
ANS-2 spec/ans-2-versioned-naming.md ANSName URI form, Identity Certificate URI SAN binding, mTLS
ANS-3 spec/ans-3-dns-publication.md DNS publication, record styles, DANE, anchor-conditional emission
ANS-4 spec/ans-4-transparency.md SCITT statements and receipts, Transparency Log, checkpoints and tiles, HCS anchoring
ANS-5 spec/ans-5-integrity-monitoring.md VerificationWorker, integrity reporting
ANS-6 spec/ans-6-agent-authentication.md Agent-to-agent authentication: badge/SCITT verification tiers, mTLS and DPoP (RFC 9449) caller methods

Worked examples for ANS-1 live at spec/examples/ans-1-examples.md. DESIGN.md is a pointer to the layered specs and to the IETF draft.

MAESTRO.md — Security analysis. Applies the CSA
MAESTRO
threat modeling framework to the ANS architecture. Maps threats and mitigations
across all seven layers, from foundation models through the agent ecosystem,
using a concrete multi-agent travel-booking scenario.

Related

trust-index-spec —
Trust evaluation. How a Trust Index crawls sealed data from federated
Registration Authorities, combines it with external signals, and scores agents
across five dimensions: integrity, identity, solvency, behavior, and safety.
Trust Manifest schema, credential formats, identity grades, federation, and
security considerations. Maintained in its own repository; not an ANS protocol
layer.

A temporary TRUST_INDEX_SPEC.md copy is maintained here
while the existing PRs #46,
#47,
#48,
#49, and
#50 are reviewed and resolved.
Accepted changes will be incorporated into trust-index-spec before this local
copy is removed.

SENDER_VERIFICATION_SPEC.md — Per-sender
email verification. JWS header-only signing with key bindings sealed in a
SCITT transparency log, discoverable via DNS. Targets business email
compromise where domain-level authentication (DKIM/SPF/DMARC) passes but the
sender is not who they claim. Not an ANS protocol feature.

API

The OpenAPI contracts and conformance schemas live under api/:

SDKs

Origin

This architecture builds on "Agent Name Service for Secure AI Agent Discovery"
by Narajala, Huang, Habler, and Sheriff (OWASP), contributed to the IETF as
an internet-draft (draft-narajala-ans).

License

Apache 2.0. See LICENSE.

Yorumlar (0)

Sonuc bulunamadi