manage-bans

Pass

Audited by Gen Agent Trust Hub on Jul 10, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • Local Script Execution: The skill utilizes a local TypeScript script (ban-ast/scripts/validate-rule.mts) to verify the syntax and effectiveness of Tree-sitter queries. This is a standard validation pattern for configuration tools, ensuring that rules are correctly formed before being applied.
  • Configuration Management: The skill manages rules stored in BANNED_AST.md files and a global file located at ~/.BANNED_AST.md. Accessing and modifying these files is the primary function of the skill, allowing for centralized and hierarchical policy enforcement across different project subtrees.
  • Policy Enforcement Mechanism: The skill describes a preToolUse hook that monitors file edits against the defined AST patterns. This mechanism provides an automated way to prevent specific code constructs, such as the use of eval(), from entering the codebase.
  • User-Defined Overrides: The implementation includes a justification override system (// <rule-name> justification: ...). This provides flexibility for developers to bypass bans when necessary, provided they document the reason directly in the source code.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 10, 2026, 02:50 PM
Security Audit — agent-trust-hub — manage-bans