Skip to main content

Frequently Asked Questions (FAQ)

Common questions about the marketplace and how to get started with AI-powered best practices.

Q: What is this marketplace for?​

It automates best practices across the software development lifecycle. Use it to:

  • Generate and maintain changelogs, READMEs, and documentation
  • Create meeting agendas and summaries with action items
  • Set up project governance (codes of conduct, contributing guidelines, issue templates)
  • Implement security audits and vulnerability scanning
  • Add licensing and intellectual property documentation
  • Configure testing frameworks and continuous integration
  • Connect to external services (GitHub, Slack, Jira) via MCP servers
  • Create custom workflows and skill templates
  • and more...

It is ideal for standardizing practices across projects, onboarding teams faster, and modernizing codebases without manual implementation.

Q: What do I need to use these best practices?​

You're recommended to have an agentic AI tool installed, such as:

  • Claude Code — command-line AI agent by Anthropic
  • Cursor — AI-powered code editor
  • Windsurf — agentic IDE
  • Aider — AI pair programming in your terminal
  • Other AI assistants with skill / MCP support

Important: the AI tool needs to be an agentic tool — not a browser-based chatbot. It should be able to read/write files, execute commands, and automate workflows.

Using an individual best practice: browse the marketplace, select a best practice, and click Install to see installation instructions. After installing, use it by following the example usage provided.

Combining best practices: the real power comes from combining multiple best practices in a single request. Your AI agent intelligently invokes the appropriate tools and skills — e.g. "Set up my project with a README, changelog, and MIT license."

Q: What's the difference between Skills, Agents, and MCP Servers?​

Skills are specialized instruction sets that guide AI agents through specific tasks — "recipes" for workflows like creating changelogs or generating docs.

Agents are autonomous systems that execute complex, multi-step workflows independently, making decisions and orchestrating entire processes.

MCP Servers (Model Context Protocol) are integrations that connect AI agents to external services and APIs — GitHub, Slack, Jira, databases, and more.

These artifacts can be combined: a skill might depend on an MCP server, or an agent might leverage multiple skills.

Q: How do I run my own marketplace?​

This site is built on a reusable, open-source framework. To run your own:

  1. Fork or clone the repository.
  2. Customize docusaurus.config.js (title, URLs, branding) and the marketplace block in static/data/registry.json.
  3. Add your best practices — see Submit a Best Practice.
  4. Deploy for free with GitHub Pages.

You can also federate other marketplaces by adding their registry URLs to marketplaceConfig.registries in docusaurus.config.js.

Q: How do I contribute my own best practice?​

  1. Review the guidelines: see Submit a Best Practice.
  2. Create your artifact: develop your skill, agent, or MCP server.
  3. Test thoroughly.
  4. Submit a pull request adding it to static/marketplace/ and an entry to static/data/registry.json.

Q: Can I use these best practices with my CI/CD pipeline?​

Yes. Best practices are designed to complement existing workflows — they can be integrated into GitHub Actions, run as pre-commit hooks, executed in automated testing pipelines, and incorporated into release processes.

Still have questions?​

Open an issue or start a discussion on the project's GitHub repository.