What problem does it solve? Documents written for agents often fail to produce consistent behavior: weak descriptions never trigger, bloated files bury key steps, and vague completion criteria let agents stop early. This Skill provides a disciplined framework for writing any agent-consumed document so the agent follows a predictable process on every run. ## Core Features & Use Cases - Context pointer design: Write descriptions and AGENTS.md lines with front-loaded leading words and one trigger per branch so material is reached reliably. - Information hierarchy and progressive disclosure: Place steps and reference on the right tier, disclose branch-specific material behind pointers, and co-locate related rules to fight sprawl. - Completion criteria and splitting: Define clear, demanding done-conditions to prevent premature completion, and split documents by sequence or invocation when the cut earns it. - Use Case: When authoring a new skill or editing CLAUDE.md, apply the pruning rules (single source of truth, no-op detection, relevance checks) to keep the document short, live, and behavior-changing. ## Quick Start Ask the agent to review your draft SKILL.md or AGENTS.md using the writing-for-agents guidance and tighten its description, steps, and completion criteria.
