skills/google/skills/google-cloud-solution-agentic-analytics-spark-knowledge-catalog/Gen Agent Trust Hub
google-cloud-solution-agentic-analytics-spark-knowledge-catalog
Pass
Audited by Gen Agent Trust Hub on Jul 17, 2026
Risk Level: SAFE
Full Analysis
- Automated Script Execution: In Phase 3, the skill generates and executes shell commands using tools like
curlandgcloudto validate the proposed architecture. This is a core feature for verifying deployments but requires the user to review and approve the generated commands before they are run. - Workspace Modification: Phase 4 involves the skill writing documentation and code files directly into the user's local workspace. This behavior is clearly disclosed and requires explicit user permission before any files are created or modified.
- Indirect Prompt Injection Surface: The skill processes user-provided business and technical requirements to generate architectural designs and validation scripts. While this is necessary for a solution design agent, it represents an entry point for untrusted data. The skill mitigates this by using structured workflows and requiring user confirmation at each phase of the generation process.
- External Resource Connectivity: The skill connects to the Google Developer Knowledge MCP server and references repositories under the
googleandgemini-cli-extensionsorganizations. These connections are used to ground the AI's recommendations in official documentation and established technical frameworks.
Audit Metadata