analyzing-database-indexes

Pass

Audited by Gen Agent Trust Hub on Apr 13, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill correctly uses scoped database CLI tools (psql, mysql, mongosh) to perform performance analysis by querying database system catalogs.
  • [COMMAND_EXECUTION]: The included script scripts/analyze_indexes.py performs generic file system directory traversal and calculates metadata statistics (size, type), though this implementation does not align with the script's name and documentation regarding database index analysis.
  • [PROMPT_INJECTION]: The skill features an indirect prompt injection surface as it processes external data from database query results.
  • Ingestion points: Data returned from database queries (table names, query statistics) enter the agent context via CLI tools.
  • Boundary markers: No specific delimiters or boundary markers are defined in SKILL.md to isolate external database content from instructions.
  • Capability inventory: The skill possesses file write (Write), edit (Edit), and shell execution (Bash) capabilities across all files.
  • Sanitization: The instructions do not define sanitization or validation protocols for processing external database content.
  • [SAFE]: No evidence of credential exposure, unauthorized data exfiltration, or persistence mechanisms was found in the provided files.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 13, 2026, 09:29 PM