Skip to content

Security: vyana-dev/open-agent-commerce

Security

SECURITY.md

Security policy

These protocols govern agent authorization, credential issuance, and payment — security reports are taken seriously.

Reporting a vulnerability

Do not open a public issue for vulnerabilities. Email security@vyana.ai with:

  • a description and impact,
  • steps to reproduce (or a proof of concept),
  • affected component (spec section, @vyana/open-agent-commerce version, or schema).

We aim to acknowledge within 3 business days and to agree a disclosure timeline with you. Coordinated disclosure is appreciated; we credit reporters unless you prefer to remain anonymous.

Scope

In scope: signature/canonicalization flaws, mandate/scope/amount bypass, replay, provenance tampering, credential-capsule encryption issues, and adapter mis-mappings that could let a request escape its intended authorization.

Out of scope: issues in upstream protocols themselves (Visa TAP, Google AP2, Stripe/OpenAI ACP) — report those to their respective stewards.

Cryptography notes

  • Signatures are Ed25519 over the canonical form of an object. Any change to canonicalization is a security-relevant, wire-breaking change.
  • UAP adapters normalize shape only and set identity.verified = false; verification is the verifier's responsibility. Treating an unverified request as verified is the canonical misuse to guard against.

There aren't any published security advisories