What problem does it solve? AI coding agents hallucinate APIs, ignore project conventions, and degrade in quality during long sessions because they receive too little, too much, or stale context. This Skill provides a structured method for curating exactly what the agent sees at each stage of a session. ## Core Features & Use Cases - Context Hierarchy Setup: Create rules files (CLAUDE.md, .cursorrules, AGENTS.md) covering tech stack, commands, conventions, and boundaries, then layer specs, source files, and error output on top. - Context Budget Management: Trim failed attempts and verbose tool output starting at 75% window capacity, compress rather than delete, and position task-critical content last to counter the lost-in-the-middle effect. - Confusion Management: Surface spec-versus-code conflicts and missing requirements as explicit options instead of silently guessing, using inline planning patterns before multi-step execution. - Use Case: When starting a new feature in a large codebase, load only the relevant spec section, the files to modify, one existing pattern example, and a project map summary so the agent follows conventions instead of inventing them. ## Quick Start Set up a CLAUDE.md rules file for my project and show me how to structure context for my next coding session.
