Skip to content

FAQ

What is this tutorial about?

It teaches practical AI agent engineering: Skills, MCP, OpenSpec, Harness Engineering, context engineering, retrieval, evaluation, and safe release workflows.

Where should I start?

Start with Quick Start. It gives you a small working loop before the longer conceptual chapters.

Why include both OpenSpec and Harness Engineering?

OpenSpec helps define what should be built. Harness Engineering helps prove that the behavior still works after prompts, tools, models, or code change.

What does MCP add?

MCP gives an agent a structured way to discover tools, resources, prompts, and context from external systems. It is useful when a workflow needs reliable boundaries instead of ad hoc prompt text.

Do I need every chapter before building?

No. Use the learning map by role. Builders can start with Skills and MCP. Maintainers should add OpenSpec, Harnesses, and evaluation gates early.

How do I contribute?

Read Contributing, run npm run docs:build, run npm run docs:check-links, and include the learning objective, setup steps, practice task, and troubleshooting notes for new tutorials.

The site previously exposed English reference links before the pages existed. The current workflow includes an internal-link audit so official navigation cannot ship with missing first-party routes.

Next: Commands or Troubleshooting.

基于 MIT 许可发布