graphql-api-design
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [DYNAMIC_EXECUTION]: The pagination.js example demonstrates a generic query builder that constructs SQL statements via string interpolation for table names (${table}) and column keys (${key}). While the example uses parameterized queries for values, it lacks validation or allow-listing for these identifiers, which is a common vector for SQL injection if input is not strictly controlled by the developer.
- [INDIRECT_PROMPT_INJECTION]: The skill enables an agent to generate functional code based on user input. The provided templates do not include explicit boundaries or sanitization logic for SQL identifiers, creating a surface where an agent might generate vulnerable code if the user provides malicious entity or field names. \n\n1. Ingestion points: Arguments passed to paginatedQuery in pagination.js (specifically table and where object keys). \n2. Boundary markers: Absent; there are no instructions to validate or restrict identifier names. \n3. Capability inventory: Code generation and template interpolation for database access. \n4. Sanitization: Present for query parameters (values), but absent for structural SQL identifiers (table and column names).
Audit Metadata