Skills
Navigate
skills/orcaqubits/agentic-commerce-skills-plugins/ucp-embedded-checkout/Gen Agent Trust Hub

ucp-embedded-checkout

Pass

Audited by Gen Agent Trust Hub on Sep 17, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill recommends fetching technical specifications from Google's official merchant documentation and the UCP developer portal. These are recognized and official sources for the described protocol.
  • [INDIRECT_PROMPT_INJECTION]: The skill describes a protocol for receiving data from an external merchant-hosted UI.
  • Ingestion points: JSON-RPC 2.0 messages received via the postMessage API from a merchant's continue_url (SKILL.md).
  • Boundary markers: Communication is restricted to the structured JSON-RPC 2.0 format.
  • Capability inventory: The platform (host) handles specific methods for acquiring payment credentials, updating line items, and completing orders.
  • Sanitization: The skill mandates strict origin validation for all messages, the use of CSP frame-ancestors to prevent unauthorized framing, and the application of the sandbox attribute (including credentialless) for the checkout iframe.
  • [SAFE]: No malicious patterns, command injections, or obfuscated contents were detected. The skill focuses on secure protocol implementation guidelines.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 17, 2026, 06:33 PM