How do I make Claude Code skills activate automatically?

Copy the skill-activation-prompt and post-tool-use-tracker hooks from the showcase, then configure skill-rules.json with trigger patterns for your project. This system analyzes prompts and file context to suggest skills, as detailed in the auto-activation breakthrough section.

Is this better than building Claude Code infrastructure from scratch?

Yes, for teams using TypeScript/React, as it saves months of trial-and-error with production-tested patterns. However, for other tech stacks, building from scratch might be more efficient due to the heavy customization needed here.

Can I use these patterns with plain JavaScript instead of TypeScript?

Partially, but the skills and examples are optimized for TypeScript, with specific guidelines for typing and tooling. You'll need to adapt the patterns, which may reduce out-of-the-box effectiveness.

How to customize skill-rules.json for my project?

Edit the JSON file to match your file paths and trigger keywords based on your domain. The README provides examples, but it requires manual tuning to align with your codebase and common prompts.

What are the essential hooks to start with in Claude Code?

Begin with skill-activation-prompt and post-tool-use-tracker hooks for auto-activation, as they work out of the box. Skip optional Stop hooks like tsc-check.sh until you customize them for your monorepo structure.

Does this work for solo developers or only teams?

It benefits both, but solo developers on small projects might find the overhead unnecessary. The infrastructure shines in team settings where consistent patterns and knowledge preservation across context resets are critical.

Open-Awesome

Claude Code Infrastructure Showcase

MITTypeScript

A reference library of production-tested Claude Code infrastructure patterns for auto-activating skills, modular skill design, and specialized agents.

GitHub

10.0k stars1.2k forks0 contributors

What is Claude Code Infrastructure Showcase?

Claude Code Infrastructure Showcase is a reference library of production-tested patterns and systems for configuring and extending Claude Code, an AI-powered development environment. It specifically solves the problem of skills not activating automatically and provides scalable infrastructure for enterprise development by offering auto-activation hooks, modular skill patterns, specialized agents, and a documentation system that survives context resets. The patterns are extracted from six months of real-world use managing a complex TypeScript microservices project.

Target Audience

Developers and teams using Claude Code who want to automate skill activation, manage complex projects, and scale their AI-assisted development workflows, particularly those working on TypeScript/JavaScript microservices or React frontends.

Value Proposition

Developers choose this over generic examples because it provides battle-tested infrastructure validated across 50,000+ lines of TypeScript in six microservices, solving real workflow problems like skill discoverability. Its unique auto-activation system via hooks and skill-rules.json ensures skills suggest themselves based on user prompts and file context, unlike standard Claude Code where skills must be manually invoked.

Overview

Examples of my Claude Code infrastructure with skill auto-activation, hooks, and agents

Use Cases

Best For

Automating Claude Code skill activation so skills suggest themselves based on context instead of requiring manual invocation.
Managing large Claude Code skills that exceed context limits using the modular 500-line rule with progressive disclosure.
Teams developing TypeScript microservices or React applications who want production-tested guardrails and patterns for backend/frontend development.
Preserving project knowledge across Claude context resets using the three-file dev docs system with the /dev-docs slash command.
Streamlining complex development tasks with specialized agents for code architecture review, refactoring, documentation generation, and error debugging.
Quickly integrating proven Claude Code infrastructure into existing projects with a 15-30 minute setup time using the provided hooks and patterns.

Not Ideal For

Teams using AI coding assistants other than Claude Code, such as GitHub Copilot or Cursor, as the infrastructure is specifically tailored to Claude Code's hook system.
Projects requiring fully functional, out-of-the-box applications rather than reference patterns to copy and adapt.
Developers working primarily with programming languages or frameworks outside TypeScript and React, since the skills and examples are heavily biased towards that stack.
Monorepos with significantly different directory structures than the example, as hooks like tsc-check.sh need extensive customization.

Pros & Cons

Pros

Auto-Activation Breakthrough

Solves the core problem of skills not activating automatically via a UserPromptSubmit hook and skill-rules.json configuration, ensuring skills suggest themselves based on user prompts and file context.

Production-Validated Patterns

Extracted from 6 months of real-world use across 50,000+ lines of TypeScript in six microservices, providing battle-tested reliability for enterprise development.

Modular Skill Architecture

Implements a 500-line rule with progressive disclosure, keeping skills within Claude's context limits—exemplified by skills like backend-dev-guidelines with 12 resource files for incremental loading.

Comprehensive Component Library

Includes 5 skills, 6 hooks, 10 agents, and 3 slash commands, all ready for integration in 15-30 minutes with clear guides like CLAUDE_INTEGRATION_GUIDE.md.

Cons

Heavy Customization Required

The settings.json and hooks like tsc-check.sh are example-only and need significant adjustments for different project structures, as admitted in the 'What Won't Work As-Is' section.

Technology Stack Lock-in

Skills and examples are heavily biased towards TypeScript microservices and React frontends, with limited utility for other languages or frameworks without adaptation.

Reference-Only Limitations

It's not a working application but a library of patterns, requiring manual copying and adaptation, which adds initial setup effort and potential integration errors.

Frequently Asked Questions

Related Projects

Superpowers

An agentic skills framework & software development methodology that works.

Stars260,162

Forks23,194

Last commit22 hours ago

Everything Claude Code

The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.

Stars232,622

Forks35,453

Last commit19 hours ago

Learn Claude Code

Bash is all you need - A nano claude code–like 「agent harness」, built from 0 to 1

Stars72,092

Forks11,687

Last commit28 days ago

Claude Scientific Skills

Turn any AI agent into an AI Scientist. The #1 Agent Skills library for science, used by 160,000+ scientists worldwide. 148 ready-to-use skills plus 100+ scientific databases covering biology, chemistry, medicine, and drug discovery. Compatible with Cursor, Claude Code, Codex, Pi, Antigravity, and the open Agent Skills standard.

Stars31,588

Forks3,150

Last commit1 day ago

Community-curated · Updated weekly · 100% open source

Found a gem we're missing?

Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.

Submit a project Star on GitHub