retail-assortment-gap-analysis
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes a local helper script
scripts/crawlora.shto communicate with the Crawlora API. - The script implements robust security measures, including mandatory path validation and a strict whitelist of permitted API routes for Shopify, Target, and IKEA.
- It securely manages the
CRAWLORA_API_KEYby writing it to a temporary curl configuration file with restricted permissions (chmod 600), ensuring the key does not appear in process lists or logs. - [INDIRECT_PROMPT_INJECTION]: The skill processes retail data (titles, descriptions, categories) fetched from external platforms like Shopify, Target, and IKEA.
- Ingestion points: Data enters the agent context as JSON responses from the
api.crawlora.netendpoints. - Boundary markers: The prompt instructions do not specify the use of delimiters when interpreting the retrieved product data.
- Capability inventory: The skill's active capabilities are restricted to making GET requests to the defined vendor API via the included bash script.
- Sanitization: The helper script validates all input paths and uses
--data-urlencodefor query parameters to prevent injection into the HTTP request structure.
Audit Metadata