What problem does it solve? AI coding agents produce inconsistent or hallucinated output when they receive too little context, too much context, or stale context. This Skill provides a structured methodology for curating what an agent sees at each stage of a session so output follows project conventions instead of inventing APIs. ## Core Features & Use Cases - Five-Level Context Hierarchy: Structures context from persistent rules files (CLAUDE.md, .cursorrules, AGENTS.md) down through specs, source files, error output, and conversation history. - Context Budget Management: Trims failed attempts and verbose tool output starting at 75% window capacity, compresses exploration into summaries, and positions task-critical content last to counter the lost-in-the-middle effect. - Confusion Management: Surfaces spec-versus-code conflicts and missing requirements as explicit options instead of silently guessing, with an inline planning pattern for multi-step tasks. - Use Case: When an agent starts ignoring your naming conventions mid-project, apply this Skill to write a rules file, restart the session at a task boundary, and reload only the relevant spec section and source files. ## Quick Start Ask the agent to audit the current session context and create a CLAUDE.md rules file covering the tech stack, commands, conventions, and boundaries for this project.
