google-cloud-solution-hybrid-search-alloydb
Pass
Audited by Gen Agent Trust Hub on Aug 28, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- Command and Infrastructure Code Generation: The skill generates technical artifacts including AlloyDB DDL, SQL scripts, Terraform manifests, and gcloud CLI commands. The agent is also instructed to execute dry-run validation commands after obtaining user approval. Users are encouraged to review all generated code before final application to their environment.
- Input Processing and Architectural Influence: The skill ingests user-provided functional and non-functional requirements to generate tailored technical designs. This represents a surface for indirect prompt injection where external input influences the agent's output. The skill manages this through a structured multi-phase workflow with mandatory approval checkpoints.
- Integration with External Information Sources: The skill retrieves design patterns and documentation from trusted Google-managed repositories and official MCP servers. These references are used to ground the agent's recommendations in current cloud best practices.
Audit Metadata