python-database-ops
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill documents patterns for handling untrusted data from web requests and databases, which creates a potential surface for indirect prompt injection if those data sources are later used in agent prompts.
- Ingestion points: Data entering through FastAPI endpoint parameters and bulk data ingestion functions.
- Boundary markers: None explicitly shown in documentation examples.
- Capability inventory: The skill is configured with Read, Write, and Bash permissions and includes code for performing database write operations.
- Sanitization: The provided code examples correctly use SQLAlchemy's parameter binding and ORM mapping, which are industry-standard protections against SQL injection.
Audit Metadata