Algorithmic Command Enforcement

What problem does it solve?

Imperative instructions (e.g., "You MUST do X") often lead to low compliance in AI agents due to rationalization. This Skill provides a meta-pattern for writing boolean decision trees that enforce 100% compliance, even under pressure, eliminating ambiguity and ensuring consistent execution.

Core Features & Use Cases

• Boolean Decision Trees: Replaces vague imperatives with clear YES/NO conditions for deterministic execution.
• Explicit Invalid Conditions: Lists common rationalizations as explicitly invalid, preventing agents from using them to bypass rules.
• Deterministic Execution Path: Uses STOP commands to ensure a single, non-negotiable path through the workflow.
• Use Case: When designing a critical workflow like TDD enforcement, use this Skill's pattern to ensure agents always write tests first and delete untested code, eliminating bypasses due to "simplicity" or "sunk cost."

Quick Start

When designing a new command or agent workflow:

1. Define a "Decision Algorithm" with Step N: Check: [Boolean condition]? -> YES/NO

2. Include a "Recovery Algorithm" for when things go wrong.

3. List "INVALID conditions" to preempt rationalizations.

4. Add a "Self-Test" section to verify comprehension.