What problem does it solve? Verbal rules shared during a session are easily lost or duplicated across documents. This Skill turns a user's spoken rule into a durable knowledge base entry, deciding whether to create a new topic or merge into an existing one, and keeps routing indexes in sync. ## Core Features & Use Cases - Rule normalization: Converts free-form dictated text into a structured rule unit with constraints, scope, and applicability. - Merge vs. create decision: Scans existing topics in .Knowledge/manifest-routing.json and index.md, then proposes merging into an existing topic, creating a new one, or splitting cross-topic rules—always with user confirmation before writing. - Routing synchronization: Updates topicPaths, topicDependencies, taskToTopicRules, matcher files, and the topic index so the new rule is discoverable by future task routing. - Use Case: A user says "all user-facing error messages must start with a verb." The Skill finds no suitable host topic, proposes creating topics/error-message-style.md, writes it after confirmation, and registers routing metadata. ## Quick Start Tell the agent to run f2s-kb-addRules and dictate the rule you want recorded, for example: "Remember this rule in the knowledge base: commit messages must start with an emoji."
