integration-tests
Warn
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADS
Full Analysis
- [EXTERNAL_DOWNLOADS] (MEDIUM): The skill instructs the user to fetch content from
https://fullstackrecipes.com/api/recipes/integration-tests. This domain is not part of the trusted external sources list. This creates a risk where the fetched content could contain malicious code or instructions.\n- [DATA_EXFILTRATION] (LOW): The skill utilizescurlfor network operations to a non-whitelisted domain. While the specific command provided does not exfiltrate sensitive files, it establishes a pattern of connecting to untrusted external infrastructure.\n- [PROMPT_INJECTION] (LOW): The skill exhibits an indirect prompt injection surface by ingesting external data (Category 8).\n - Ingestion points: Data fetched from
fullstackrecipes.com.\n - Boundary markers: None provided to separate the external recipe from system instructions.\n
- Capability inventory: Network access (
curl).\n - Sanitization: None specified for the external content.
Audit Metadata