mcpfoldMCP glossary
Plain, accurate definitions of the core Model Context Protocol concepts — what an MCP server is, what a client is, how tools and the context window relate, and how mcpfold fits in. Each page links down to the product, the directory, and the docs.
- MCP server
- An MCP server is a program that exposes tools, resources, and prompts to AI applications over the Model Context Protocol, so any compatible client can use its capabilities.
- Model Context Protocol
- The Model Context Protocol (MCP) is an open standard for connecting AI applications to external tools and data sources through a common client–server interface.
- MCP client
- An MCP client is the part of an AI application — an editor, chat app, or agent — that connects to MCP servers and lets the model use their tools.
- MCP tools
- MCP tools are the individual actions an MCP server exposes to a model — each with a name, a description, and a JSON-Schema definition of its inputs.
- Context window
- A context window is the maximum amount of text, measured in tokens, that a language model can consider at once — including every MCP tool schema loaded into it.
- Secret reference
- A secret reference is a placeholder such as ${env:GITHUB_TOKEN} that points to a secret stored elsewhere, so the raw value is never written into a config file.
- MCP config manager
- An MCP config manager is a tool that maintains your MCP servers in one source of truth and applies that configuration across every client, instead of hand-editing a separate file per application.
Ready to use it? Install mcpfold or browse the MCP server directory.