common-workflow-writing

Token-efficient workflows require strict structure, frontmatter with name and description, line limits, and self-check validation to enforce minimal, actionable content without prose explanations. Merge sequential steps and standardize required sections.

Common anti-patterns in SKILL.md writing include using pre-filled data rows, adding prose explanations, and skipping automated self-check prompts. Avoiding these prevents token bloat and eliminates unnecessary optimization passes during automation.

Create a standard workflow template by enforcing template-driven design over pre-filled data, using mandatory frontmatter, defining required sections, and setting line limits so runtime generation can fill specifics with crisp, executable steps.

AI workflows consume too many tokens due to unoptimized structure, pre-filled rows, and excessive prose. Imposing strict size limits, merging sequential steps, and enforcing automated self-check prompts ensure minimal, runnable artifacts.

Yes, frontmatter with name and description is a mandatory requirement for SKILL.md generation. It enforces concise, token-efficient workflow writing by standardizing structure and preventing unnecessary optimization passes during automation.

Strict line limits in workflow templates restrict verbose explanations and pre-filled data, requiring crisp, executable steps. This constraint prevents token bloat but means all content must be minimal and actionable without prose context.