What problem does it solve?
Documentation in large repositories often becomes duplicated, bloated, or hard to discover. This Skill enforces consistent guidelines when creating or updating docs so content stays minimal, canonical, and discoverable by both humans and AI agents.
Core Features & Use Cases
- Structured doc workflow: Follows a five-step process from understanding the topic to validating the result.
- Frontmatter standards: Defines conventions for
name, description, allowed-tools, and invocation flags so skills and agents are tool-discoverable.
- Quality principles: Applies progressive disclosure, docs locality, and canonicalization rules, with validation steps using
rg and wc -l.
- Use Case: When adding a new guide under
docs/ or updating AGENTS.md in the ledger-live monorepo, use this Skill to decide the right location, keep the file under roughly 100 lines, and avoid duplicating existing knowledge.
Quick Start
Ask the agent to create or update a documentation file in this repo while following the docs-update-or-create guidelines.