What problem does it solve?
Agent-facing documents like SKILL.md files and AGENTS.md degrade over time: descriptions fail to trigger, content duplicates across files, and stale lines accumulate. This Skill provides a disciplined model for writing documents that AI agents consume, covering context pointers, information hierarchy, completion criteria, leading words, and pruning so every line changes agent behavior.
Core Features & Use Cases
- Context pointer design: Write skill descriptions and AGENTS.md lines that front-load trigger words, name distinct branches, and carry a Covers clause so agents reliably reach the right material.
- Information hierarchy and progressive disclosure: Decide what stays inline as steps or reference versus what moves behind a pointer into disclosed reference files, preventing document sprawl.
- Skill mechanics for Stacks: Covers frontmatter validation rules, model-invoked versus user-invoked skills, the app/Skills override model, and what buddy setup:ai materializes.
- Use Case: When adding a new skill under app/Skills/ or editing a bundled one in storage/framework/defaults/ai/skills, apply this Skill to write a compliant frontmatter, structure the body, and prune no-op sentences before committing.
Quick Start
Use the stacks-writing-for-agents skill to review and rewrite my new SKILL.md so its description triggers reliably and the body follows the information hierarchy.
