> For the complete documentation index, see [llms.txt](https://docs.mozilla.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.mozilla.ai/mcpd/integrations/sdks.md).

# SDKs

`mcpd` exposes an HTTP REST API, so any language that can make an HTTP request can drive it. The SDKs on this page remove the boilerplate.

There are two kinds, and they answer different questions:

| Kind       | Question it answers                            | Runs in                |
| ---------- | ---------------------------------------------- | ---------------------- |
| Client SDK | How does my application call the tools?        | Your agent application |
| Plugin SDK | How do I intercept what passes through `mcpd`? | A plugin process       |

***

## Client SDKs

A client SDK lists the servers `mcpd` manages, reads tool schemas, and calls tools from your own code.

| Language              | Package            | Repository                                                               | Install                        |
| --------------------- | ------------------ | ------------------------------------------------------------------------ | ------------------------------ |
| Python                | `mcpd`             | [mcpd-sdk-python](https://github.com/mozilla-ai/mcpd-sdk-python)         | `uv add mcpd`                  |
| JavaScript/TypeScript | `@mozilla-ai/mcpd` | [mcpd-sdk-javascript](https://github.com/mozilla-ai/mcpd-sdk-javascript) | `npm install @mozilla-ai/mcpd` |

Both follow the same shape. You construct a client against the daemon address, then reach a tool through the server that provides it:

```python
from mcpd import McpdClient

client = McpdClient(api_endpoint="http://localhost:8090")
result = client.call.time.get_current_time(timezone="UTC")
```

```typescript
import { McpdClient } from "@mozilla-ai/mcpd";

const client = new McpdClient({ apiEndpoint: "http://localhost:8090" });
const result = await client.servers.time.tools.get_current_time({ timezone: "UTC" });
```

Each SDK also hands a framework its tool definitions in one call, so you do not write a conversion layer. The Python SDK targets [any-agent](https://github.com/mozilla-ai/any-agent), and the JavaScript SDK targets LangChain JS and the Vercel AI SDK.

{% hint style="info" %}
**The endpoint is the daemon's API address**

`mcpd daemon` binds `0.0.0.0:8090` unless `api.addr` says otherwise, so `http://localhost:8090` works for a daemon on the same machine. See [Daemon Configuration](/mcpd/configuration/daemon-configuration.md) when the daemon runs elsewhere, and [Execution Context](/mcpd/configuration/execution-context.md) for API key handling.
{% endhint %}

***

## Plugin SDKs

A plugin SDK wraps the gRPC stubs generated from [`mcpd-proto`](https://github.com/mozilla-ai/mcpd-proto) and the startup handshake, so a plugin only has to implement the behavior you care about. [Writing a Plugin](/mcpd/configuration/writing-a-plugin.md) covers the protocol itself.

| Language | Package                      | Repository                                                                       | Install                                            |
| -------- | ---------------------------- | -------------------------------------------------------------------------------- | -------------------------------------------------- |
| .NET     | `MozillaAI.Mcpd.Plugins.Sdk` | [mcpd-plugins-sdk-dotnet](https://github.com/mozilla-ai/mcpd-plugins-sdk-dotnet) | `dotnet add package MozillaAI.Mcpd.Plugins.Sdk`    |
| Go       | `mcpd-plugins-sdk-go`        | [mcpd-plugins-sdk-go](https://github.com/mozilla-ai/mcpd-plugins-sdk-go)         | `go get github.com/mozilla-ai/mcpd-plugins-sdk-go` |
| Python   | `mcpd-plugins`               | [mcpd-plugins-sdk-python](https://github.com/mozilla-ai/mcpd-plugins-sdk-python) | `pip install mcpd-plugins`                         |
| Rust     | `mcpd-plugins-sdk`           | [mcpd-plugins-sdk-rust](https://github.com/mozilla-ai/mcpd-plugins-sdk-rust)     | `cargo add mcpd-plugins-sdk`                       |

The Python, Rust, and .NET repositories each include an `examples/` directory with runnable plugins.

### Community plugin SDKs

These are maintained by their authors. Mozilla AI does not maintain or test them.

| Language | Repository                                                                | Maintainer                                 |
| -------- | ------------------------------------------------------------------------- | ------------------------------------------ |
| PHP      | [mcpd-plugins-sdk-php](https://github.com/shoemoney/mcpd-plugins-sdk-php) | [@shoemoney](https://github.com/shoemoney) |

***

## No SDK for your language

Write against the API directly. The [API Reference](/mcpd/reference/api-reference.md) documents every endpoint, and `mcpd daemon` serves the same specification at `/docs`.

To build a plugin instead, generate stubs from [`mcpd-proto`](https://github.com/mozilla-ai/mcpd-proto) with your language's gRPC tooling and implement the handshake described in [Writing a Plugin](/mcpd/configuration/writing-a-plugin.md).


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.mozilla.ai/mcpd/integrations/sdks.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.