What problem does it solve? Writing skills that agents actually trigger and load correctly is hard: vague descriptions never match, bloated bodies waste context, and duplicated scopes drift apart. This Skill provides the authoritative authoring pattern for any skill in .agents/skills/, covering discovery mechanics, description writing, progressive disclosure, and anti-drift maintenance. ## Core Features & Use Cases - Trigger-optimized descriptions: Rules for the 1024-character ceiling, the 300–500 character comfort zone, front-loaded triggers, and action-verb vs. topic-vocabulary anchoring. - Progressive disclosure structure: Guidance on the three loading levels (metadata, body, resources) and when to move detail into references/ or templates into assets/. - Scope and maintenance governance: Criteria for skill vs. always-on instruction, one-skill-per-scope splitting and merging, forbidden body sections, and the anti-drift maintenance block. - Use Case: When asked to create a new domain skill for a project, apply this pattern to write a valid frontmatter, a lean prescriptive body, and ready-to-copy templates from assets/. ## Quick Start Use the skill-authoring skill to review this SKILL.md and fix its frontmatter, description, and structure.
