ap2-human-not-present-flow

Warn

Audited by Snyk on Mar 31, 2026

Risk Level: MEDIUM
Full Analysis

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

  • Third-party content exposure detected (high risk: 0.90). The SKILL.md "Fetch live docs" steps explicitly require fetching public third‑party pages (e.g., https://ap2-protocol.org/specification/ and web-searching site:github.com for google-agentic-commerce samples), which the agent must read to determine Intent Mandate fields and flow behavior, so untrusted web content could materially influence its decisions.

MEDIUM W009: Direct money access capability detected (payment gateways, crypto, banking).

  • Direct money access detected (high risk: 1.00). The skill is explicitly about implementing an autonomous purchasing flow ("autonomous agent purchasing", "human-not-present transaction flow") that includes creating/signed Intent Mandates and a Phase ("If within bounds → SA proceeds with payment" and "Payment processing (similar to human-present Phase 4-6)"). Although it does not name a specific payment gateway, its primary and explicit purpose is to enable the agent to decide to and execute payments on behalf of the user. That constitutes direct financial execution authority.

Issues (2)

W011
MEDIUM

Third-party content exposure detected (indirect prompt injection risk).

W009
MEDIUM

Direct money access capability detected (payment gateways, crypto, banking).

Audit Metadata
Risk Level
MEDIUM
Analyzed
Mar 31, 2026, 06:09 AM
Issues
2