Summary

This repository offers a curated reference library of production-tested Claude Code infrastructure patterns. It addresses critical challenges such as automatic skill activation and effective context management, providing developers with reusable components to scale Claude Code for enterprise development. The primary benefit is accelerated integration of robust, battle-tested solutions.

How It Works

The core innovation is automatic skill activation, driven by a UserPromptSubmit hook that analyzes user prompts and file context against skill-rules.json to suggest or trigger relevant skills. Skills are modularized using a "500-line rule" and progressive disclosure, breaking down complex functionalities into a main file and smaller, context-efficient resource files. This design prevents context window limitations and ensures skills are readily available when needed.

Quick Start & Requirements

This is a reference library, not a deployable application; users must copy desired components into their projects. Integration involves selecting and copying essential hooks (e.g., skill-activation-prompt, post-tool-use-tracker), updating settings.json (which requires customization for specific project structures, service names, and server configurations), and adding skills by copying their directories and configuring skill-rules.json. Estimated integration time for basic setup is 15-30 minutes. Key resources include CLAUDE_INTEGRATION_GUIDE.md, skills documentation, hooks setup guides, and agent guides.

Highlighted Details

• Auto-activating Skills: Implements a UserPromptSubmit hook and skill-rules.json configuration for automatic skill suggestion and activation based on user prompts and file context.
• Modular Skill Design: Employs a "500-line rule" with progressive disclosure, structuring skills into a main file and multiple smaller resource files to manage Claude's context window effectively.
• Specialized Agents: Offers 10 pre-built agents for complex tasks such as code architecture review, refactoring assistance, documentation generation, and error debugging.
• Persistent Dev Docs: Provides a three-file pattern ([task]-plan.md, [task]-context.md, [task]-tasks.md) for generating and preserving project context across context resets, accessible via /dev-docs commands.
• Production Proven: Patterns are derived from real-world use across 6 microservices, over 50,000 lines of TypeScript, and 6 months of daily Claude Code application.

Maintenance & Community

The repository is presented as a personal showcase of production-tested patterns. Community engagement is encouraged through starring the repo, reporting issues, and contributing examples. Specific community channels like Discord or Slack are not detailed.

Licensing & Compatibility

The project is released under the MIT License, permitting free use in both commercial and personal projects without significant restrictions.

Limitations & Caveats

This repository serves as a reference library; components must be manually copied and integrated. The provided settings.json is illustrative and requires significant customization for specific monorepo structures, service names, and server environments. Optional hooks like tsc-check.sh demand substantial adaptation. Example domains are generic and need to be tailored to specific project logic.