What problem does it solve? Writing OpenCode command definitions without clear standards leads to bloated files, misplaced logic, and inconsistent frontmatter. This Skill provides concrete quality criteria so commands stay thin, delegating detailed judgment tables to skills and deterministic processing to scripts. ## Core Features & Use Cases - Frontmatter and Structure Rules: Enforces a description-only frontmatter, integer-only top-level steps, and runtime path conventions (.opencode/...) instead of source paths. - Size and Responsibility Thresholds: Defines line count targets (100 lines ideal, 150 operational limit), step counts (5-12), token budgets, and clear rules for what belongs in a command versus a skill, reference, or script. - Review Checklists and DoD: Supplies a Definition of Done and review checklist covering delegation contracts, subagent edit safety, common policy identifiers (POL-), and writing quality. - Use Case: When creating a new /agentdev/* command or refactoring an oversized one, load this Skill to decide what to extract into orchestration skills or scripts and verify the result against the checklist. ## Quick Start Review my new OpenCode command definition against the command authoring standards and tell me what to extract into skills or scripts.
