managing-database-tests
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSDYNAMIC_EXECUTIONCREDENTIALS_UNSAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill downloads a utility script from the official Anthropics GitHub repository (
raw.githubusercontent.com/anthropics/skills). Per security policy, this is a trusted organization and the download is considered safe. - [DYNAMIC_EXECUTION]: The provided
generate_test_data.pyutility contains agenerate_scriptfunction that writes shell scripts to the file system and programmatically sets executable permissions (chmod 755) on them. This represents a dynamic code generation pattern. - [CREDENTIALS_UNSAFE]: The
assets/docker-compose.ymlfile contains hardcoded passwords for PostgreSQL, MySQL, and MongoDB services (e.g.,test_password,root_password). While standard for local test isolation, these are hardcoded secrets. - [INDIRECT_PROMPT_INJECTION]: The skill's primary function involves reading and processing user-defined database schemas and test data templates (JSON/SQL). This creates an attack surface where instructions embedded in processed data could potentially influence the AI agent's behavior during the testing lifecycle.
Audit Metadata