Skip to content

Commit f8b70fe

Browse files
dvdksndvdksn
authored
Merge pull request #23601 from dvdksn/mcp-docs-remake
mcp: rewrite mcp toolkit/gateway documentation
2 parents 86d5f13 + 1a49fa7 commit f8b70fe

File tree

8 files changed

+220
-364
lines changed

8 files changed

+220
-364
lines changed

‎content/manuals/ai/mcp-catalog-and-toolkit/_index.md‎

Lines changed: 57 additions & 67 deletions
Original file line numberDiff line numberDiff line change
@@ -11,7 +11,7 @@ weight: 30
1111
description: Learn about Docker's MCP catalog on Docker Hub
1212
keywords: Docker, ai, mcp servers, ai agents, extension, docker desktop, llm, docker hub
1313
grid:
14-
- title: Get started
14+
- title: Get started with MCP Toolkit
1515
description: Learn how to quickly install and use the MCP Toolkit to set up servers and clients.
1616
icon: explore
1717
link: /ai/mcp-catalog-and-toolkit/get-started/
@@ -23,77 +23,67 @@ grid:
2323
description: Learn about the MCP Toolkit to manage MCP servers and clients
2424
icon: /icons/toolkit.svg
2525
link: /ai/mcp-catalog-and-toolkit/toolkit/
26+
- title: MCP Gateway
27+
description: Learn about the underlying technology that powers the MCP Toolkit
28+
icon: developer_board
29+
link: /ai/mcp-catalog-and-toolkit/toolkit/
30+
- title: Docker Hub MCP server
31+
description: Explore about the Docker Hub server for searching images, managing repositories, and more
32+
icon: device_hub
33+
link: /ai/mcp-catalog-and-toolkit/toolkit/
2634
---
2735

2836
{{< summary-bar feature_name="Docker MCP Catalog and Toolkit" >}}
2937

3038
[Model Context Protocol](https://modelcontextprotocol.io/introduction) (MCP) is
31-
an open protocol that standardizes how AI applications access external tools and data sources. Through a client-server architecture, applications such as [Gordon](/manuals/ai/gordon/_index.md) or Claude Desktop act as clients that send requests to MCP servers, which then process these requests and deliver the necessary context to AI models.
32-
33-
Building and managing MCP tools can be complex. Docker MCP Catalog and Toolkit simplifies this by providing a secure, streamlined way to build, share, and run MCP tools, addressing common challenges developers face.
34-
35-
Docker MCP Catalog and Toolkit simplifies the developer experience across the following areas:
36-
37-
- Discovery: Centralized discovery through a verified catalog of versioned tools.
38-
- Credential management: Secure credential management with OAuth-based authentication.
39-
- Execution: Safe execution by running tools in isolated, containerized environments.
40-
- Portability: True portability. MCP tools work seamlessly across Claude, Cursor, Visual Studio Code, and other platforms without requiring code changes.
41-
42-
This integrated approach means you can launch MCP servers in seconds, add new tools through either CLI or GUI interfaces, and rely on Docker's pull-based infrastructure for trusted delivery of your tools.
43-
44-
![MCP overview](./images/mcp-overview.svg)
45-
46-
To understand how Docker MCP Catalog and Toolkit achieves this streamlined experience, let's explore the underlying architecture and how its key components work together.
47-
48-
## MCP architecture overview
49-
50-
MCP follows a client-server architecture that enables standardized communication between AI applications and external tools. This foundational understanding helps you build and integrate MCP solutions effectively.
51-
52-
### MCP servers
53-
54-
MCP servers are specialized programs that provide specific tools and capabilities to AI models through [Model Context Protocol](https://modelcontextprotocol.io/introduction). Each server focuses on a particular domain, like GitHub integration, database access, or file management. You can think of them as extensions that give AI models the ability to interact with external systems and data sources.
55-
56-
### MCP clients
57-
58-
MCP clients are the bridge between AI applications and MCP servers. They enable AI systems to discover, connect to, and interact with external tools and data sources through the standardized MCP protocol. A single client can connect to multiple MCP servers, while each server can also serve multiple clients, creating a flexible many-to-many relationship that allows for rich integration possibilities.
59-
60-
When you interact with an AI application that supports MCP, the client establishes connections to configured servers, discovers available tools and resources from all connected servers, executes requests when needed, and manages contextual information to enhance AI responses.
61-
62-
### MCP Gateway
63-
64-
The [MCP Gateway](/manuals/ai/mcp-catalog-and-toolkit/mcp-gateway.md) is Docker's open-source solution which connects MCP servers to MCP clients. It provides a unified endpoint that consolidates multiple MCP servers from the Docker MCP Catalog into a single, manageable connection point. Instead of clients connecting directly to individual MCP servers, the MCP Gateway sits in between, providing centralized connectivity, secure layer, and enterprise controls.
65-
66-
In the context of Docker's MCP ecosystem:
67-
68-
- MCP Catalog hosts the MCP servers (the tools/capabilities)
69-
- MCP Gateway orchestrates and secures those servers
70-
- MCP clients connect through the Gateway to access cataloged servers
71-
72-
## How MCP components work together
73-
74-
MCP components communicate through a structured flow that connects your development environment to external services. The following diagram shows how MCP works in practice with the key components:
75-
76-
1. The MCP client (VS Code) serves as your development environment where you interact with AI assistance.
77-
2. The MCP Gateway acts as a secure orchestration layer, managing and routing connection between the client (VS Code) and the GitHub Official MCP server.
78-
3. The MCP server acts as the bridge, providing specialized capabilities. In the following example, the GitHub Official server offers tools like creating pull requests, creating branches, and retrieving issues.
79-
4. The data source (GitHub) provides the actual service or platform that the MCP server connects to for real-time information and actions.
80-
81-
This architecture enables seamless integration where you can access GitHub functionality directly from VS Code through standardized MCP communication. The server translates requests between your development environment and external services, making it easy to incorporate various tools and data sources into your AI-assisted workflow without complex custom integrations.
82-
83-
![Example of the GitHub MCP server](./images/mcp-toolkit-overview.png)
84-
85-
> [!TIP]
86-
> Example:
87-
> If you work in Visual Studio Code's _agent mode_ and ask it to create a
88-
> branch in GitHub, it needs an MCP server provided by GitHub to do that.
89-
>
90-
> The MCP server provided by GitHub provides _tools_ to your model to perform
91-
> atomic actions, like:
92-
>
93-
> - `Create a PR`
94-
> - `Create a branch`
95-
> - ...
96-
>
39+
an open protocol that standardizes how AI applications access external tools
40+
and data sources. By connecting LLMs to local development tools, databases,
41+
APIs, and other resources, MCP extends their capabilities beyond their base
42+
training.
43+
44+
Through a client-server architecture, applications such as Claude, ChatGPT, and
45+
[Gordon](/manuals/ai/gordon/_index.md) act as clients that send requests to MCP
46+
servers, which then process these requests and deliver the necessary context to
47+
AI models.
48+
49+
MCP servers extend the utility of AI applications, but running servers locally
50+
also presents several operational challenges. Typically, servers must be
51+
installed directly on your machine and configured individually for each
52+
application. Running untrusted code locally requires careful vetting, and the
53+
responsibility of keeping servers up-to-date and resolving environment
54+
conflicts falls on the user.
55+
56+
## Docker MCP features
57+
58+
Docker provides three integrated components that address the challenges of
59+
running local MCP servers:
60+
61+
MCP Catalog
62+
: A curated collection of verified MCP servers, packaged and distributed as
63+
container images via Docker Hub. All servers are versioned, come with full
64+
provenance and SBOM metadata, and are continuously maintained and updated with
65+
security patches.
66+
67+
MCP Toolkit
68+
: A graphical interface in Docker Desktop for discovering, configuring, and
69+
managing MCP servers. The Toolkit provides a unified way to search for servers,
70+
handle authentication, and connect them to AI applications.
71+
72+
MCP Gateway
73+
: The core open source component that powers the MCP Toolkit. The MCP Gateway
74+
manages MCP containers provides a unified endpoint that exposes your enabled
75+
servers to all AI applications you use.
76+
77+
This integrated approach ensures:
78+
79+
- Simplified discovery and setup of trusted MCP servers from a curated catalog
80+
of tools
81+
- Centralized configuration and authentication from within Docker Desktop
82+
- A secure, consistent execution environment by default
83+
- Improved performance since applications can share a single server runtime,
84+
compared to having to spin up duplicate servers for each application.
85+
86+
![MCP overview](./images/mcp-overview.svg)
9787

9888
## Learn more
9989

‎content/manuals/ai/mcp-catalog-and-toolkit/get-started.md‎

Lines changed: 31 additions & 26 deletions
Original file line numberDiff line numberDiff line change
@@ -8,32 +8,34 @@ weight: 10
88

99
{{< summary-bar feature_name="Docker MCP Toolkit" >}}
1010

11-
Docker Model Context Protocol (MCP) Toolkit makes it easy to set up, manage, and run containerized MCP servers, and connect them to AI agents. It provides secure defaults, one-click setup, and support for a growing ecosystem of LLM-based clients. This page shows you how to get started quickly with the Docker MCP Toolkit.
11+
The Docker MCP Toolkit makes it easy to set up, manage, and run containerized
12+
Model Context Protocol (MCP) servers, and connect them to AI agents. It
13+
provides secure defaults and support for a growing ecosystem of LLM-based
14+
clients. This page shows you how to get started quickly with the Docker MCP
15+
Toolkit.
1216

13-
## Requirements
17+
## Setup
1418

1519
Before you begin, make sure you meet the following requirements to get started with Docker MCP Toolkit.
1620

17-
### Download and install Docker Desktop
18-
19-
- Install [Docker Desktop version 4.42.0](/manuals/desktop/release-notes.md#4420) and later.
20-
- To try the MCP Toolkit Learning center walkthrough, you must install [Docker Desktop version 4.46.0](/manuals/desktop/release-notes.md#4460) and later.
21-
22-
### Enable Docker MCP Toolkit
23-
24-
1. Open the Docker Desktop settings and select **Beta features**.
25-
2. Select **Enable Docker MCP Toolkit**.
26-
3. Select **Apply**.
21+
1. Download and install the latest version of [Docker Desktop](/get-started/get-docker/).
22+
2. Open the Docker Desktop settings and select **Beta features**.
23+
3. Select **Enable Docker MCP Toolkit**.
24+
4. Select **Apply**.
2725

2826
There are multiple ways to get started with Docker MCP Toolkit. You can:
29-
- Try the [Learning center](#learning-center) walkthrough in Docker Desktop, available in [Docker Desktop version 4.46.0](/manuals/desktop/release-notes.md#4460) and later.
30-
- Alternatively, follow the step-by-step instructions on this page to use Docker Desktop or the CLI to [Install an MCP server](#install-an-mcp-server), [add a client, and test your setup](#install-an-mcp-client-and-test-your-setup) with example prompts.
3127

32-
## Learning center
28+
- The **Learning center** in Docker Desktop provides walkthroughs and resources
29+
to help you get started with Docker products and features.
3330

34-
The **Learning center** in Docker Desktop provides walkthroughs and resources to help you get started with Docker products and features.
31+
On the **MCP Toolkit** page, the **Get started** walkthrough that guides you
32+
through installing an MCP server, connecting a client, and testing your
33+
setup.
3534

36-
On the **MCP Toolkit** page, Docker Desktop shows the **Get started** walkthrough that guides you through installing an MCP server, connecting a client, and testing your setup.
35+
- Alternatively, follow the step-by-step instructions on this page to use
36+
Docker Desktop or the CLI to [Install an MCP server](#install-an-mcp-server),
37+
[add a client, and test your setup](#install-an-mcp-client-and-test-your-setup)
38+
with example prompts.
3739

3840
## Install an MCP server
3941

@@ -44,7 +46,7 @@ On the **MCP Toolkit** page, Docker Desktop shows the **Get started** walkthroug
4446
2. Search for the **GitHub Official** server from the catalog and then select the plus icon to add it.
4547
3. In the **GitHub Official** server page, select the **Configuration** tab and select **OAuth**.
4648

47-
>[!NOTE]
49+
> [!NOTE]
4850
>
4951
> The type of configuration required depends on the server you select. For the GitHub Official server, you must authenticate using OAuth.
5052
@@ -68,7 +70,7 @@ On the **MCP Toolkit** page, Docker Desktop shows the **Get started** walkthroug
6870
$ docker mcp oauth authorize github
6971
```
7072

71-
>[!NOTE]
73+
> [!NOTE]
7274
>
7375
> The type of configuration required depends on the server you select. For the GitHub Official server, you must authenticate using OAuth.
7476
@@ -79,11 +81,11 @@ On the **MCP Toolkit** page, Docker Desktop shows the **Get started** walkthroug
7981
```console
8082
$ docker mcp server enable playwright
8183
```
82-
{{< /tab >}}
83-
{{< /tabs >}}
8484

85-
You’ve now successfully added an MCP server. Next, install an MCP client and test your setup with an example prompt.
85+
{{< /tab >}}
86+
{{< /tabs >}}
8687

88+
You’ve now successfully added an MCP server. Next, install an MCP client and test your setup with an example prompt.
8789

8890
## Install an MCP client and test your setup
8991

@@ -96,17 +98,18 @@ After you've installed MCP servers, you can add clients to the MCP Toolkit. Thes
9698
2. Find **Claude Desktop** and select **Connect**.
9799

98100
If Claude Desktop isn't installed, select **Download** to install it, then select **Connect**.
101+
99102
3. Restart Claude Desktop if it's running, and it can now access all the servers in the MCP Toolkit.
100-
4. Open Claude Desktop and run a test by submitting the following prompt using the Sonnet Opus 4.1 model:
103+
4. Open Claude Desktop and run a test by submitting the following prompt:
101104

102105
```text
103106
Take a screenshot of the header element on docs.docker.com
104107
```
108+
105109
Claude prompts you for permissions and shares a screenshot of the header element from the Docker documentation page.
106110

107111
![Screenshot showing the header of Docker docs.](./images/claude-desktop-example.png)
108112

109-
110113
{{< /tab >}}
111114
{{< tab name="From the Docker CLI">}}
112115

@@ -115,21 +118,23 @@ After you've installed MCP servers, you can add clients to the MCP Toolkit. Thes
115118
```console
116119
$ docker mcp client connect claude-desktop --global
117120
```
121+
118122
2. Restart Claude Desktop if it's running, and it can now access all the servers in the MCP Toolkit.
119123
3. Open Claude Desktop and run a test by submitting the following prompt using the Sonnet 4 model:
120124

121125
```text
122126
Take a screenshot of the header element on docs.docker.com
123127
```
128+
124129
Claude prompts you for permissions and shares a screenshot of the header element from the Docker documentation page.
125130

126131
![Screenshot showing the header of Docker docs.](./images/claude-desktop-example.png)
127132

128133
{{< /tab >}}
129134
{{< /tabs >}}
130135

131-
## Related pages
136+
## Further reading
132137

133138
- [MCP Toolkit](/manuals/ai/mcp-catalog-and-toolkit/toolkit.md)
134139
- [MCP Catalog](/manuals/ai/mcp-catalog-and-toolkit/catalog.md)
135-
- [Open-source MCP Gateway](/manuals/ai/mcp-catalog-and-toolkit/mcp-gateway.md)
140+
- [MCP Gateway](/manuals/ai/mcp-catalog-and-toolkit/mcp-gateway.md)

0 commit comments

Comments
 (0)