SQL Query Rewriter
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides purely instructional content for SQL optimization. It includes safety-conscious advice, such as using
EXPLAINwithoutANALYZEif a query cannot be run safely, and ensuring equivalence proofs before deployment. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process user-provided SQL queries and database EXPLAIN plans. While this constitutes an attack surface for untrusted data, the skill's lack of high-privilege tool access or network operations restricts the potential impact to the generation of incorrect SQL suggestions.
- Ingestion points: User-provided query text and
EXPLAIN (ANALYZE)output described in the 'Workflow' section ofSKILL.md. - Boundary markers: None explicitly defined in the prompt logic to isolate user query data from instruction logic.
- Capability inventory: The skill generates text and SQL code; no autonomous tool usage or file system writes are defined in the frontmatter.
- Sanitization: No specific sanitization or validation of the input SQL syntax is performed before rewriting.
Audit Metadata