blink-database
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The
blink_db_querytool allows for the execution of arbitrary SQL commands against the database, which includes data modification (INSERT/UPDATE/DELETE) and schema manipulation (CREATE/DROP TABLE). - [INDIRECT_PROMPT_INJECTION]: The skill processes data retrieved from database tables, which creates a vulnerability where malicious instructions stored in the data could influence the agent's behavior.
- Ingestion points: Data returned by the
blink_db_querytool and standard CRUD operations such aslist,get, andcount(SKILL.md). - Boundary markers: No specific delimiters or instructions are provided to help the agent differentiate between database content and its own instructions.
- Capability inventory: The agent has full access to query, create, update, and delete database records, which could be exploited if malicious content is processed.
- Sanitization: There is no evidence of input validation or content sanitization for data retrieved from the database before it is provided to the agent context.
Audit Metadata