sql-optimizer

Pass

Audited by Gen Agent Trust Hub on Apr 12, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: The skill is entirely composed of documentation and instructional content. It does not contain any executable code (scripts, binaries, or commands) that could be used for malicious purposes.
  • [NO_CODE]: There are no script files or automated tasks included in the skill. It functions as a knowledge base to guide the agent's reasoning during SQL performance reviews.
  • [SAFE]: No network requests, file exfiltration patterns, or hardcoded credentials were detected. The skill only processes user-provided SQL queries to offer optimization advice.
  • [SAFE]: No prompt injection or behavior override attempts were found within the instructions or metadata. The skill adheres to its stated purpose of database performance analysis.
  • [SAFE]: Analysis of the input surfaces (SQL queries and EXPLAIN plans) reveals a low risk of indirect prompt injection, as the skill's capabilities are restricted to generating textual advice and SQL suggestions for the user, with no automated execution of the generated code.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 12, 2026, 06:25 AM
Security Audit — agent-trust-hub — sql-optimizer