glossary · The process

MCP server

definition

An MCP server is a program that exposes tools, resources and prompts to AI applications through the Model Context Protocol, so any MCP-compatible client can use its capabilities without a custom integration.

in one line: Exposes tools and data to any AI client that speaks MCP.

in context

Where does an MCP server fit?

One server. Many clients. In ConvOps, the server owns the domain, and any MCP client can use it.

client connectsone URL
sign inOAuth
list toolstyped inputs
model calls one
result back
two ways to say it

What is an MCP server, in plain words?

Same idea at two depths: the plain version, then what the engine actually does.

A server describes what it can do. Your AI client shows those abilities to the model.

Write it once, and it works from every compatible client.

for engineers

Tools, resources and prompts. Local over stdio, or remote over HTTP.

ConvOps is a remote server at https://mcp.convops.app/ with OAuth, and API keys for automation.

the longer answer

Why does an MCP server matter?

The Model Context Protocol, introduced by Anthropic as an open standard in November 2024, splits an AI integration into two sides. The client lives inside the AI application and speaks to the model. The server owns a domain, such as a database, a code host, a design tool or a work system, and describes what it can do in a form the client understands. Because both sides follow one protocol, one server works with many clients.

A server offers three kinds of capability. Tools are actions the model can call, with a name, a description and a typed input schema. Resources are data the client can read, addressed by URI. Prompts are reusable templates a user can pick. Servers run either locally, started by the client as a process and spoken to over standard input and output, or remotely, reached over HTTP. Remote servers usually sign users in with OAuth, so the server knows whose data each call touches.

For the people using an AI client, a good MCP server feels like a built-in ability. For the people building one, the hard part is not the protocol but the design: which actions to expose, how to describe them so the model uses them well, and how to keep authority with the signed-in user rather than with whatever the model types.

in convops

How does an MCP server work in ConvOps?

ConvOps is a remote MCP server at https://mcp.convops.app/. Adding it is one line of client configuration; on first use the client opens a browser to sign in, and the server identifies you and your workspace from that sign-in. API keys exist for automation and clients without OAuth. Its tools cover tasks, workflows, gates, memory, pools, schedules, executors and governance, so the whole operations layer appears inside the AI client you already use. ConvOps sells no models: the client keeps its own.

questions

What do people ask about an MCP server?

What does an MCP server do?

An MCP server, such as ConvOps, exposes a system's capabilities to AI clients in a standard form: tools the model can call, resources the client can read and prompts a user can pick. ConvOps exposes its tasks, workflows, gates, memory, pools, schedules, executors and governance this way, so they appear inside the AI client you already use. Any MCP client can use a server without a custom integration.

What is the difference between a local and a remote MCP server?

ConvOps is a remote MCP server, which shows the difference in practice. A local server is started by the client as a process and talks over standard input and output, while a remote server is reached over HTTP, usually with OAuth sign-in. Because ConvOps is remote, it knows whose data each call touches from the sign-in. API keys cover automation and clients without OAuth.

How do I add the ConvOps MCP server?

Add https://mcp.convops.app/ to your client's MCP settings to connect ConvOps. In Claude Code the command is claude mcp add --transport http convops https://mcp.convops.app/. On first use the client opens a browser to sign in, and the server identifies you and your workspace from that sign-in.

Who should connect an MCP server to their AI client?

Anyone who wants an AI client to work with a system it cannot reach today. Connecting the ConvOps MCP server suits team members in Claude Code, Cursor, ChatGPT or Codex who need shared tasks, workflows and memory inside the client they already use, with one sign-in and no custom integration.

What can an AI client do through the ConvOps MCP server?

Through the ConvOps MCP server, an AI client works with tasks, workflows, gates, memory, work pools, schedules, executors and governance as tools in the conversation. The client asks for the current step of a task, does the work and asks to advance, and the server checks that step's gates before the cursor moves. Each change is recorded with the agent's label and the signed-in account behind it.

What makes an MCP server well designed?

ConvOps treats the protocol as the solved part of an MCP server and the design as the hard one. A good server exposes clear actions, describes each one so the model uses it well, and keeps authority with the signed-in user rather than with whatever the model types. In ConvOps, the user recorded on every change comes from the verified sign-in or API key, and no tool parameter can set it.