What problem does it solve? Documentation and code comments often become stale or misleading because they contain time-dependent phrases, historical references, duplicated information, or restate what the code already shows. This Skill provides a checklist-driven workflow that keeps Markdown documents and code comments accurate and understandable for readers months later. ## Core Features & Use Cases - Four-phase writing workflow: Identify the reader and purpose before writing, apply Evergreen/Why-First/SSOT/Reader-Friendly principles while writing, then run three self-check rounds after writing. - Time-dependency detection: Grep commands catch phrases like "previously", "currently", "planned", PR numbers, commit SHAs, and personal names so they can be rewritten as timeless fact statements. - SSOT enforcement: Checks for duplicated information across files with grep and converts duplicates into references to a single source of truth. - Use Case: When asked to update a README, add a rule to .claude/rules/, write JSDoc for an exported function, or draft a PR description, run this checklist before and after editing to avoid documentation lint violations. ## Quick Start Use the evergreen-writing skill to review and rewrite this README section so it contains no time-dependent expressions or duplicated content.
