google-merchant-feed
Fail
Audited by Gen Agent Trust Hub on Sep 24, 2026
Risk Level: CRITICALEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill recommends the use of standard development libraries and references official Google documentation.
- Suggests installation of
xmlbuilder2via npm for Node.js implementations. - References official Google specifications from
base.google.com,google.com, andsupport.google.com. - Contains a placeholder URL
https://your-frontend-domain.comwhich triggered an automated phishing alert; however, this domain is used strictly as a documentation example for the user to substitute with their actual frontend URL. - [INDIRECT_PROMPT_INJECTION]: The skill defines patterns for processing product data into XML feeds, representing a potential injection surface.
- Ingestion points: Data is retrieved from back-end databases (e.g., Django QuerySets, Mongoose models, or CMS APIs).
- Boundary markers: The skill uses structured XML tags as natural delimiters for data fields.
- Capability inventory: The skill generates XML responses for syndication to Google Merchant Center.
- Sanitization: All implementation examples provided in the skill include mandatory XML escaping. The Python version uses the
xml.etree.ElementTreebuilt-in escaping, the Node.js version usesxmlbuilder2's automatic escaping, and the Next.js example includes a customescapeXmlutility function to sanitize strings.
Recommendations
- Contains 1 malicious URL(s) - DO NOT USE
Audit Metadata