### Zread MCP SSE Endpoint Example

Source: https://zread.ai/mcp/index

Example URL for accessing Zread MCP via Server-Sent Events (SSE). This method adheres to the official MCP specification and is recommended for most IDEs and clients that support MCP.

```bash
https://api.z.ai/api/mcp/zread/sse?Authorization=zai_api_key
```

--------------------------------

### Search Documentation Tool (search_doc)

Source: https://zread.ai/mcp/index

The `search_doc` tool facilitates searches within a GitHub repository's documentation to understand its purpose, setup, core modules, and workflows. It can also retrieve information on related news, issues, commits, and contributors.

--------------------------------

### Zread MCP Streamable HTTP Endpoint Example

Source: https://zread.ai/mcp/index

Example URL for accessing Zread MCP via Streamable HTTP. This method is adapted for clients that support HTTP streaming MCP, such as some self-built Agents or OpenAI compatible implementations.

```bash
https://api.z.ai/api/mcp/zread/mcp
```

--------------------------------

### Add Zread MCP Server via Claude Code CLI (Bash)

Source: https://zread.ai/mcp/index

Use this command in the Claude Code CLI to add the Zread MCP server to your configuration. Replace `zai_api_key` with your actual API key.

```bash
claude mcp add -s user -t http zread https://api.z.ai/api/mcp/zread/mcp --header "Authorization: Bearer zai_api_key"
```

--------------------------------

### Available Zread MCP Tools

Source: https://zread.ai/mcp/index

Zread MCP offers three atomic tools for code understanding and retrieval: `search_doc`, `get_repo_structure`, and `read_file`.

```APIDOC
## Zread MCP Tools

### `search_doc`

#### Description
Performs searches within the documentation of a GitHub repository to help understand repository purpose, setup, core modules, workflows, and related news or contributors.

#### Parameters
- **repo_url** (string) - Required - The URL of the GitHub repository.
- **query** (string) - Required - The search query string.

### `get_repo_structure`

#### Description
Views the structure of a GitHub repository or subdirectory, allowing the model to understand project hierarchy and module organization.

#### Parameters
- **repo_url** (string) - Required - The URL of the GitHub repository.
- **path** (string) - Optional - The subdirectory path within the repository to explore.

### `read_file`

#### Description
Reads the full content of a single file in a GitHub repository, used for reviewing implementation details, code analysis, or refactoring suggestions.

#### Parameters
- **repo_url** (string) - Required - The URL of the GitHub repository.
- **file_path** (string) - Required - The path to the file within the repository.
```

--------------------------------

### Zread MCP SSE Endpoint

Source: https://zread.ai/mcp/index

The SSE (Server-Sent Events) endpoint adheres to the official MCP recommended specification and is suitable for most IDEs/Clients that support MCP. This is the recommended method for priority use.

```APIDOC
## Zread MCP SSE Endpoint

### Description
This endpoint allows clients to communicate with Zread MCP using Server-Sent Events, following the official MCP specification. It's ideal for IDEs and clients that support MCP.

### Method
POST (or other methods as supported by MCP clients)

### Endpoint
`/api/mcp/zread/sse`

### Query Parameters
- **Authorization** (string) - Required - Your ZAI API key for authentication.

### Request Body
Refer to the MCP specification for the exact request body format, which typically includes the tool to be called and its parameters.

### Request Example
```json
{
  "tool_code": "search_doc",
  "parameters": {
    "repo_url": "https://github.com/example/repo",
    "query": "how to install"
  }
}
```

### Response
#### Success Response (200)
The response will be a stream of Server-Sent Events containing the results from the requested tool.

#### Response Example
```
event: tool_code
data: {"tool_code": "search_doc"}

event: message
data: {"content": "Installation instructions can be found in the README file. Run `npm install`."}

```
```

--------------------------------

### Zread MCP Server Configuration

Source: https://zread.ai/mcp/index

Configuration for the Zread MCP server, specifying the type, URL, and authorization headers for the 'zread' service.

```json
{
  "mcpServers": {
    "zread": {
      "type": "streamableHttp",
      "url": "https://api.z.ai/api/mcp/zread/mcp",
      "headers": {
        "Authorization": "Bearer zai_api_key"
      }
    }
  }
}
```

--------------------------------

### Zread MCP Streamable HTTP Endpoint

Source: https://zread.ai/mcp/index

The Streamable HTTP endpoint is adapted for clients that support HTTP streaming MCP, such as some self-built Agents or OpenAI compatible implementations.

```APIDOC
## Zread MCP Streamable HTTP Endpoint

### Description
This endpoint provides Zread MCP capabilities via HTTP streaming, suitable for custom agents or OpenAI compatible clients.

### Method
POST (or other methods as supported by MCP clients)

### Endpoint
`/api/mcp/zread/mcp`

### Headers
- **Authorization** (string) - Required - Your ZAI API key for authentication.

### Request Body
Refer to the MCP specification for the exact request body format, which typically includes the tool to be called and its parameters.

### Request Example
```json
{
  "tool_code": "get_repo_structure",
  "parameters": {
    "repo_url": "https://github.com/example/repo"
  }
}
```

### Response
#### Success Response (200)
The response will be a stream of data chunks representing the results from the requested tool.

#### Response Example
```json
{
  "tool_code": "get_repo_structure",
  "output": "{\"dir_name\": \"repo\", \"children\": [{\"dir_name\": \"src\"}, {\"file_name\": \"README.md\"}]}"
}
```
```

=== COMPLETE CONTENT === This response contains all available snippets from this library. No additional content exists. Do not make further requests.