database-design

Pass

Audited by Gen Agent Trust Hub on Oct 27, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill includes a bash utility scripts/lint-migration.sh used to scan SQL migration files for common safety issues such as unintended table drops or column renames. The script utilizes standard Unix tools and includes validation logic to reject target paths containing shell metacharacters.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external SQL data which creates a potential surface for indirect prompt injection attacks.
  • Ingestion points: User-provided SQL migration files, schema definitions, and raw query strings analyzed by the agent using the skill's instructions or the lint-migration.sh script.
  • Boundary markers: Absent. There are no explicit instructions or delimiters defined to prevent the agent from obeying instructions that might be embedded within the analyzed SQL comments or strings.
  • Capability inventory: The skill provides a shell script (scripts/lint-migration.sh) capable of performing directory traversal and file reading using find and grep.
  • Sanitization: The lint-migration.sh script performs basic sanitization by checking the input path against a regex of shell metacharacters ($, `, ;, |, &, () before execution.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 27, 2026, 05:38 PM