skill-name

AI agent documentation with progressive disclosure and anti-pattern warnings is created using a structured SKILL.md template with YAML frontmatter. This template provides sections for decision frameworks and separates reference files to ensure token-efficient loading of examples and edge-case handling.

The best way to structure expert-level rules for AI assistants is using a templated SKILL.md format with YAML frontmatter. This approach reduces ambiguity and context usage by separating core guidance from detailed reference files, enabling token-efficient loading when agents consume the guidance.

You build a new AI skill guide by copying a structured template and editing the SKILL.md fields with your domain's specific rules and examples. Optional AGENTS.md files, reference directories, and script or asset directories can be added to support complex workflows and edge-case handling.

Yes, you can use separate reference files to handle edge cases in AI skill documentation. This template structure enables token-efficient loading of examples and edge-case handling, allowing AI agents to consume detailed workflows without overwhelming the core guidance context.

No specific dependencies or environments are required to function, as the skill operates with zero dependencies. It requires only a structured SKILL.md file with YAML frontmatter, alongside optional AGENTS.md, reference files, and script or asset directories to operate effectively.

You should not use a progressive disclosure template for AI agent guidance when your task lacks complex decision frameworks or anti-pattern warnings. If your domain rules are simple enough to fit entirely within a single context window without separate reference files, this structured approach may introduce unnecessary overhead.