bigtable-basics
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- Indirect Prompt Injection Surface: The skill processes user-provided inputs such as table names, row keys, and SQL queries which are interpolated into CLI commands and API calls. This creates a surface for potential input manipulation. 1. Ingestion points: Data entering through placeholders in references/cli_data_access.md, references/infrastructure_management.md, and references/sql_guide.md. 2. Boundary markers: No explicit delimiters or warnings to ignore embedded instructions are provided in the command templates. 3. Capability inventory: Commands include gcloud and cbt operations for creating, reading, updating, and deleting Bigtable instances and tables. 4. Sanitization: No sanitization or validation logic is specified for the interpolated variables.
- CLI Configuration Management: The skill provides instructions for initializing the ~/.cbtrc configuration file for the cbt tool. While this involves writing to the local file system, it is a standard procedure for configuring Bigtable data access.
- Usage Attribution and Telemetry: The skill requests that specific identifiers be included in command environments and User-Agent headers for telemetry. These identifiers are static and do not contain user-specific or sensitive data.
- Cloud API Interaction: The skill demonstrates how to query the Dataplex API using standard authentication tokens. These operations target official vendor endpoints and are consistent with the skill's purpose of resource management.
Audit Metadata