anti-rogue-actions

Prevent destructive API calls by enforcing pre-execution sanity checks that validate permissions, parameter bounds, and business logic guardrails before operations run, stopping unauthorized or nonsensical actions in production environments.

Pre-execution sanity checks are validation steps that enforce permission boundaries, parameter bounds, and scope narrowing before an API call executes, ensuring destructive operations require explicit confirmation and target the smallest possible scope.

Yes, you need PowerShell installed to execute the validation scripts. You copy the sanity-check script to your project's .ai_scripts directory and run it with command and target size flags to validate operations.

Yes, dry-run verification allows you to check operations for destructive commands and target size before actual execution. Running the sanity-check script validates parameter limits and scope without committing changes.

Enforce parameter bounds by ensuring quantity, depth, scope, and limit parameters are explicitly set and validated before execution. Scope narrowing restricts operations to the smallest possible target area to prevent runaway actions.

The best way to confirm destructive chains is to validate each step individually, enforcing business logic guardrails and permission boundaries at every stage, requiring explicit confirmation before proceeding to the next destructive action.