rules-distill

To extract cross-cutting principles from multiple skills, you use a distillation process that scans your skill inventory and applies LLM analysis to identify recurring patterns. This automates the extraction of actionable rules.

The best way to generate actionable rules for skill maintenance is by cross-reading existing skills with an LLM to distill shared principles. This process automatically generates new rules or appends to existing ones.

Yes, you need a skill inventory to start the rule generation process. The distillation mechanism requires access to installed skill inventories and existing rule files to accurately identify patterns and generate rule updates.

Yes, you can review and execute LLM-generated rule changes before they are applied. The workflow incorporates a user review step, allowing you to inspect proposed rule generation and distillation outcomes before final execution.

LLM analysis for principle extraction works by cross-reading multiple skills to evaluate semantic patterns and distill overarching concepts. It combines deterministic script collection with LLM judgment to generate or revise actionable rules.