SMART POLE Chat Enforcer Skill

This "Chat Enforcer" version of SMART POLE is designed for chat-agent workflows. It acts as a mandatory gatekeeper that will not release control until requirements are clearly defined.

Scope: Prompt refinement and <master_prompt> handoff only.
Not suitable for coding-agent execution workflows (file edits, terminal commands, test execution).

Key Differences

• Output: Terminates with a <master_prompt> XML block for automated parsing.
• Workflow: Designed to be Step 1 in a chat-agent chain.
• Thinking: Uses <thinking> blocks (in Claude) to ensure rigorous analysis before output.

How to use this Skill

1. Configure: Set this skill as the entry point for your agent.
2. Execute: The agent will interview the user.
3. Capture: Watch for the XML block to trigger the next step.

The 9 Categories (Weighted Scoring)

Locale Conditional Core: If Aim is domain-sensitive (legal, finance, healthcare, HR, cross-border, culture-sensitive) → Locale becomes 🔴 Core (weight 2.0). Otherwise → 🟡 Contextualizer (weight 1.5).

Example Conditional Weight (v3.1): For Deterministic tasks (code, algorithms) → Example = 🟡 1.5. For Advisory → 1.0. For Generative → 🟢 0.5.

Threshold: ≥ 67% + All Core categories confirmed