Resources reside in external systems, and applications need a way to connect to and query them. MCP provides common communication rules, services offer specific capabilities and resources, and applications decide how to use the results.
First, Let's Look at Who's Doing What
Applications manage tasks and connections; client components handle communication with services; services provide tools, resources, or prompt templates. The model generating answers is another layer—MCP itself is not a model.
Understanding Connection, Invocation, and Completion Separately
The course materials in this site experiment actually discovered the tool get_course_material, and querying silk-road-01 returned materials for 7th grade, 40 minutes, and M01–M03. This demonstrated connecting and returning results, but did not invoke the model to generate a lesson plan.
| Success Indicators | What Else Is Needed |
|---|---|
| Service connected | Confirm whether the required capabilities exist |
| Tool invocation successful | Verify whether the returned materials are the target content |
| Model has organized a lesson plan | Continue verifying facts, objectives, class duration, and deliverables |
Decide for Yourself Whether You Need to Go Deeper
When a few pieces of material can be provided directly, there's no need to connect to additional services. When you frequently query systems for updated materials, you can explore whether the application supports the corresponding methods and permissions.
To actually observe the connection process, read Usage Guide; to understand roles and information flow, read Mechanism Explanation. Foundational learning does not require running commands or entering API Keys first.
Related Concepts for This Section
- Model Context Protocol (MCP): An open protocol that enables AI applications to connect to external capabilities and resources following common rules.
- Tool: A specific capability that a system can invoke, such as searching, reading files, computing, or generating files.
- Permission: The scope of accessible resources and executable operations controlled by the application, service, and execution environment.
General definitions and technical sources can be found on the related concepts pages, verified on 2026-09-09. The examples in this section are created for teaching purposes; optional checks are provided for understanding and do not restrict continued reading.