skill-integration

Skill discovery uses keyword-based automatic activation with progressive disclosure—lightweight metadata (~50 tokens) stays in context while full content loads on-demand. This pattern scales to 100+ skills without context bloat, enabling multi-skill composition across discovery, design, testing, security, and documentation tasks.

Progressive disclosure separates skill metadata (standardized reference format) from full content (5,000–15,000 tokens). Metadata remains in context for efficient lookup; complete details load only when needed. This prevents context bloat while maintaining end-to-end activation sequencing across complex agent workflows.

Yes. Multi-skill composition combines skills across discovery, design, testing, security, and documentation workflows. Standardized skill reference formats and activation sequencing enable seamless coordination, allowing agents to sequence and compose 100+ skills efficiently within context constraints.

This Skill applies to Claude Code 2.0+ prompts and workflows. It standardizes how agents reference and activate skills through keyword-based triggers and progressive disclosure, scaling to support enterprise-grade skill libraries without exceeding token budgets.

Standardized skill reference formats define consistent patterns for how agents cite and activate skills. These formats integrate with progressive disclosure to enable automatic keyword matching and on-demand content loading, supporting reliable skill composition across multi-step workflows.