catalog-import-export
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill documentation and provided code snippets follow industry-standard security and performance patterns for ETL (Extract, Transform, Load) operations.
- [INDIRECT_PROMPT_INJECTION]: The skill defines a process for ingesting external data from CSV and XML files. While this creates a surface for indirect prompt injection, the risk is mitigated through robust defensive programming practices demonstrated in the instructions.
- Ingestion points: Data enters the system via CSV file streams in the
parseCatalogCsvfunction and the task descriptions for merchant onboarding. - Boundary markers: The skill uses structured data parsing to transform raw text into validated objects, preventing raw data from influencing logic flow directly.
- Capability inventory: The skill's suggested implementations use file system reads and database upsert capabilities, which are scoped to the specific task of catalog management.
- Sanitization: The implementation explicitly uses
zodschema validation, enforcing regex patterns for URL handles, numeric coercion for prices, and mandatory fields to ensure data integrity.
Audit Metadata