What problem does it solve? Teams repeatedly correct AI agents for the same style or convention violations because those rules are never written down. This Skill turns ad-hoc corrections and preferences from a conversation into a persistent, reusable instructions file. ## Core Features & Use Cases - Conversation Mining: Reviews the chat history to extract corrections, enforced patterns, and project-specific conventions worth persisting. - Guided Clarification: Asks targeted questions about scope, affected file types, and whether a rule is mandatory or a preference when intent is unclear. - Iterative Refinement: Drafts the instruction, probes the weakest or most ambiguous parts, and finalizes with a summary plus suggested follow-up customizations. - Use Case: After repeatedly telling the agent to "always use early returns" and "never use default exports", invoke this Skill to capture those rules in a .instructions.md file that applies automatically in future sessions. ## Quick Start Ask the agent to create an instructions file that enforces the coding conventions we discussed in this conversation.
