sql-expert

Pass

Audited by Gen Agent Trust Hub on Aug 10, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill includes scripts/sql_helper.py, which utilizes SQLAlchemy to execute SQL queries. This is a functional requirement for the skill's purpose and is implemented using standard database connection patterns.
  • [PROMPT_INJECTION]: The skill instructions and reference materials (references/best-practices.md, references/common-pitfalls.md) explicitly prioritize the use of parameterized queries to mitigate SQL injection risks, demonstrating a security-first approach to query design.
  • [EXTERNAL_DOWNLOADS]: The documentation suggests installing standard, well-known database drivers (psycopg2-binary, sqlalchemy, mysql-connector-python, pyodbc) from the official Python Package Index (PyPI).
  • [SAFE]: Indirect Prompt Injection Surface Evaluation: 1. Ingestion points: User requirements and SQL snippets provided to the agent in the context of database tasks. 2. Boundary markers: Absent. 3. Capability inventory: SQL execution via scripts/sql_helper.py. 4. Sanitization: Detailed instructions and best practices for using parameterized queries are provided to the agent to prevent exploitation of the execution capability.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 10, 2026, 03:21 PM
Security Audit — agent-trust-hub — sql-expert