writing-sql
Pass
Audited by Gen Agent Trust Hub on Sep 5, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill describes a data ingestion surface where external RDBMS records are processed via Handlebars templates (e.g.,
{{{record.fieldName}}}). While this represents a potential SQL injection surface if user-controlled data is improperly handled, the instructions provide best practices for quoting and identify that the platform uses parameterized queries. The capability inventory includes executing SQL statements and vendor-specific CLI commands (celigo metadata,celigo exports invoke) for testing. Boundary markers are discussed in the context of triple vs. double braces for escaping. The risk is consistent with the primary purpose of the skill. - [COMMAND_EXECUTION]: The skill references the
celigoCLI tool suite (e.g.,celigo metadata,celigo exports invoke) for schema discovery and query testing. These are legitimate administrative tools provided by the skill's author, Celigo, and are used within the expected developer workflow.
Audit Metadata