bigtable-basics
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it instructs the agent to process data from Bigtable and Dataplex and use it to construct CLI commands and SQL queries via string interpolation of placeholders (e.g., {table_name}, {row_key}, {search_term}).\n
- Ingestion points: Data retrieved from Bigtable via cbt commands and search terms used for Dataplex metadata (identified in references/cli_data_access.md and references/dataplex.md).\n
- Boundary markers: The skill includes a 'Safety Rule' in SKILL.md requiring explicit user confirmation before making changes to non-emulator databases, which acts as a procedural boundary.\n
- Capability inventory: The skill utilizes the gcloud and cbt CLI tools for infrastructure management and data manipulation, and performs network requests via curl to Google Cloud APIs (references/infrastructure_management.md, references/cli_data_access.md, references/dataplex.md).\n
- Sanitization: There is no evidence of input validation, escaping, or sanitization logic to ensure that data inserted into the CLI placeholders is safe.
Audit Metadata