ucp-permalink
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines a system for processing untrusted data from browser-addressable URL parameters.
- Ingestion points: Item paths and field-path query parameters described in SKILL.md.
- Boundary markers: The instructions explicitly mandate treating query strings as fully untrusted input and validating every field-path.
- Capability inventory: The skill provides guidance for implementing logic that interacts with cart and checkout states using standard agent tools.
- Sanitization: Detailed requirements are provided for same-origin validation of redirect URLs and filtering of sensitive PII or credentials from pass-through parameters.
- [EXTERNAL_DOWNLOADS]: The skill references external technical resources to ensure compliance with the UCP specification.
- Fetches the live specification via web search targeting ucp.dev.
- References a public sample implementation repository on GitHub (Universal-Commerce-Protocol/samples).
Audit Metadata