What problem does it solve? Building Claude Code extensions requires knowing many scattered configuration formats—skill frontmatter, agent definitions, plugin manifests, hook events, settings hierarchies—and getting any of them wrong silently breaks behavior. This Skill consolidates the official authoring rules into one reference so extensions are created correctly the first time. ## Core Features & Use Cases - Extension Authoring Reference: Covers Skills, sub-agents, plugins, slash commands, hooks, memory, settings, sandboxing, headless mode, and dev containers with official documentation excerpts. - Advanced Agent Patterns: Documents two-agent long-task patterns, orchestrator-worker architectures, context engineering, and tool design best practices from Anthropic engineering guidance. - Validation Guardrails: Includes rationalization tables, red flags, and a verification checklist (e.g., CLAUDE.md under 40K characters, CSV allowed-tools format, hook timeout requirements). - Use Case: When asked to create a new Claude Code skill or configure a PreToolUse hook, this Skill supplies the exact frontmatter format, directory structure, and common pitfalls to avoid. ## Quick Start Ask the agent to create a new Claude Code skill or sub-agent and it will apply the official format rules and validation checklist from this reference.
