English | 简体中文

AgentSkills

A curated collection of Claude Code / agent skills for real-world engineering and creative workflows.

About

This repository is a skills portfolio, not a single deployable application. Most top-level folders are standalone skills with:

• SKILL.md (required entrypoint)
• optional references/, scripts/, examples/, assets/

Quick Start (3 Steps)

1. Pick a skill from the Skill Index.
2. Open that skill's SKILL.md and read the trigger + workflow instructions.
3. Run any skill-specific scripts/commands in that skill context.

Skill Index (Clickable)

Category	Jump	Skills
Agent & Skill Engineering	Go	Agent Development · Skill Development · skill-creator · dispatching-parallel-agents · find-skills · skill-lookup · mcp-builder
Backend Architecture / API / Security	Go	architecture-patterns · backend-patterns · api-design-principles · auth-implementation-patterns · microservices-patterns · postgresql-table-design
Event-Driven & CQRS	Go	cqrs-implementation · event-store-design · projection-patterns · saga-orchestration
CI/CD & Delivery Quality	Go	deployment-pipeline-design · github-actions-templates · gitlab-ci-patterns · code-reviewer
Frontend Design & Testing	Go	frontend-design · ui-ux-pro-max · vue-best-practices · web-artifacts-builder · webapp-testing
Documentation & Office Automation	Go	doc-coauthoring · readme-generate · internal-comms · docx · pdf · pptx · xlsx
Visualization & Creative Design	Go	mermaid-visualizer · excalidraw-diagram · algorithmic-art · canvas-design · theme-factory · brand-guidelines · slack-gif-creator
Domain-Specific	Go	Add Admin API Endpoint

Skill Categories

1) Agent & Skill Engineering

• Agent Development
• Skill Development
• skill-creator
• dispatching-parallel-agents
• find-skills
• skill-lookup
• mcp-builder

2) Backend Architecture / API / Security

• architecture-patterns
• backend-patterns
• api-design-principles
• auth-implementation-patterns
• microservices-patterns
• postgresql-table-design

3) Event-Driven & CQRS

• cqrs-implementation
• event-store-design
• projection-patterns
• saga-orchestration

4) CI/CD & Delivery Quality

• deployment-pipeline-design
• github-actions-templates
• gitlab-ci-patterns
• code-reviewer

5) Frontend Design & Testing

• frontend-design
• ui-ux-pro-max
• vue-best-practices
• web-artifacts-builder
• webapp-testing

6) Documentation & Office Automation

• doc-coauthoring
• readme-generate
• internal-comms
• docx
• pdf
• pptx
• xlsx

7) Visualization & Creative Design

• mermaid-visualizer
• excalidraw-diagram
• algorithmic-art
• canvas-design
• theme-factory
• brand-guidelines
• slack-gif-creator

8) Domain-Specific

• Add Admin API Endpoint (Ghost Admin API)

Repository Structure

.
├── <skill-name>/
│   ├── SKILL.md
│   ├── references/   (optional)
│   ├── scripts/      (optional)
│   ├── examples/     (optional)
│   └── assets/       (optional)
├── CLAUDE.md
├── README.md
├── README.zh-CN.md
└── LICENSE

Common Commands (Skill-Specific)

There is no single root build/lint/test pipeline. Run commands in each skill context.

# Validate an agent file
bash "Agent Development/scripts/validate-agent.sh" <path-to-agent.md>

# Initialize and bundle a web artifact project
bash scripts/init-artifact.sh <project-name>
bash scripts/bundle-artifact.sh

# Run Playwright automation with local server lifecycle management
python scripts/with_server.py --server "npm run dev" --port 5173 -- python your_automation.py

# Generate README for a Spring Boot module
python readme-generate/scripts/readme_generator.py <spring-module-path>

# Skill evaluation and packaging loop
python -m scripts.aggregate_benchmark <workspace>/iteration-N --skill-name <name>
python -m scripts.run_loop --eval-set <eval.json> --skill-path <skill-path> --model <model-id> --max-iterations 5 --verbose
python -m scripts.package_skill <path/to/skill-folder>

Contributing

Contributions are welcome.

1. Fork this repository and create a feature branch.
2. Keep changes scoped to the target skill directory.
3. Update related docs/examples when workflow behavior changes.
4. Run skill-specific validation scripts where applicable.
5. Open a PR with clear summary and test/verification notes.

Roadmap

• Add lightweight metadata cards for each skill (purpose, inputs, outputs, stack).
• Add discoverability tags and cross-links between related skills.
• Add automated link checks for SKILL.md and referenced docs.
• Add contribution templates for new skill submissions.

Acknowledgements

• Claude Code and the broader agent engineering ecosystem.
• Contributors and maintainers sharing reusable workflows.

License

MIT