### Install Dependencies and Run Development Server

Source: https://github.com/simpleyyt/ai-manus/blob/main/frontend/README.md

Install project dependencies using npm and start the development server.

```bash
npm install

npm run dev
```

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

### Run Mockserver Service Individually

Source: https://github.com/simpleyyt/ai-manus/blob/main/AGENTS.md

Starts the mock LLM server for development and testing purposes. Requires Python, pip, and uvicorn. Installs dependencies from requirements.txt.

```bash
cd mockserver
pip install -r requirements.txt
uvicorn main:app --host 0.0.0.0 --port 8090 --reload
```

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

### Start Development Services

Source: https://github.com/simpleyyt/ai-manus/blob/main/AGENTS.md

Start all development services, including backend, frontend, and mockserver.

```bash
./dev.sh up -d
```

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

### Install Project Dependencies

Source: https://github.com/simpleyyt/ai-manus/blob/main/sandbox/README.md

Installs all development dependencies for the project using uv. Ensure uv is installed before running this command.

```bash
uv sync --dev
```

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

### Run Frontend Service Individually

Source: https://github.com/simpleyyt/ai-manus/blob/main/AGENTS.md

Installs frontend dependencies and starts the Vite development server. Sets the backend URL for API requests. Requires Node.js and npm.

```bash
cd frontend
npm install
BACKEND_URL=http://localhost:8000 npm run dev
```

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

### Install Project Dependencies

Source: https://github.com/simpleyyt/ai-manus/blob/main/backend/README.md

After installing uv, use this command to install all project dependencies defined in the pyproject.toml file.

```bash
uv sync
```

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

### Copy Environment Example File

Source: https://github.com/simpleyyt/ai-manus/blob/main/README.md

Copies the example environment file to a new file named '.env'. This file will be used to configure the application's environment variables.

```shell
cp .env.example .env
```

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

### Start Development Environment with Docker Compose

Source: https://github.com/simpleyyt/ai-manus/blob/main/AGENTS.md

Quickly start the development environment including frontend, backend, sandbox, mockserver, MongoDB, and Redis. Ensure to copy and edit the .env file first, setting at least the API_KEY.

```bash
cp .env.example .env
# Edit .env — at minimum set API_KEY to any non-empty string
./dev.sh up -d
```

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

### Example MCP GitHub Server Configuration

Source: https://github.com/simpleyyt/ai-manus/blob/main/docs/en/mcp.md

An example mcp.json configuration for integrating with the GitHub API, specifying the command, transport, and environment variables.

```json
{
  "mcpServers": {
    "github": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-github"
      ],
      "transport": "stdio",
      "enabled": true,
      "description": "GitHub API integration",
      "env": {
        "GITHUB_TOKEN": "your_github_token_here"
      }
    }
  }
}
```

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

### Start Development Server

Source: https://github.com/simpleyyt/ai-manus/blob/main/sandbox/README.md

Starts the FastAPI development server using uvicorn. This command reloads the server automatically on code changes.

```bash
uv run uvicorn app.main:app --host 0.0.0.0 --port 8080 --reload
```

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

### Example .env File Configuration

Source: https://github.com/simpleyyt/ai-manus/blob/main/sandbox/README.md

Provides an example of how to configure the sandbox service using environment variables in a .env file. These settings control CORS origins, service timeouts, and log levels.

```dotenv
ORIGINS=http://localhost:3000,https://example.com
SERVICE_TIMEOUT_MINUTES=60
LOG_LEVEL=DEBUG
```

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

### Install uv Package Manager

Source: https://github.com/simpleyyt/ai-manus/blob/main/backend/README.md

Use this command to install the uv package manager, a fast Python package installer and resolver.

```bash
pip install uv
```

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

### Example .env File for AI Manus Configuration

Source: https://github.com/simpleyyt/ai-manus/blob/main/docs/en/quick_start.md

This file provides example configurations for various services including models, databases, and search engines. Uncomment and modify lines to suit your environment. It's recommended for managing settings outside of docker-compose.yml.

```ini
# Model provider configuration
API_KEY=
API_BASE=http://mockserver:8090/v1

# Model configuration
MODEL_NAME=deepseek-chat
TEMPERATURE=0.7
MAX_TOKENS=2000

# MongoDB configuration
#MONGODB_URI=mongodb://mongodb:27017
#MONGODB_DATABASE=manus
#MONGODB_USERNAME=
#MONGODB_PASSWORD=

# Redis configuration
#REDIS_HOST=redis
#REDIS_PORT=6379
#REDIS_DB=0
#REDIS_PASSWORD=

# Sandbox configuration
#SANDBOX_ADDRESS=
SANDBOX_IMAGE=simpleyyt/manus-sandbox
SANDBOX_NAME_PREFIX=sandbox
SANDBOX_TTL_MINUTES=30
SANDBOX_NETWORK=manus-network
#SANDBOX_CHROME_ARGS=
#SANDBOX_HTTPS_PROXY=
#SANDBOX_HTTP_PROXY=
#SANDBOX_NO_PROXY=

# Browser engine configuration
# Options: playwright, browser_use (default)
# - playwright:   uses Playwright directly via CDP (stable, well-tested)
# - browser_use:  uses the browser_use library's BrowserSession via CDP
#                 (richer DOM state extraction via AI-friendly selector map)
#BROWSER_ENGINE=browser_use

# Search engine configuration
# Options: baidu, baidu_web, google, bing, bing_web, tavily
# baidu: uses the Baidu Qianfan AI Search API (requires BAIDU_SEARCH_API_KEY)
# baidu_web: scrapes Baidu search results with browser impersonation (no API key needed)
# bing: uses the official Bing Web Search API (requires BING_SEARCH_API_KEY)
# bing_web: scrapes Bing search results directly (no API key needed)
SEARCH_PROVIDER=bing_web

# Baidu search configuration, only used when SEARCH_PROVIDER=baidu
# Get your API key from https://console.bce.baidu.com/qianfan/ais/console/onlineService
#BAIDU_SEARCH_API_KEY=

# Bing search configuration, only used when SEARCH_PROVIDER=bing
# Get your API key from https://www.microsoft.com/en-us/bing/apis/bing-web-search-api
#BING_SEARCH_API_KEY=

# Google search configuration, only used when SEARCH_PROVIDER=google
#GOOGLE_SEARCH_API_KEY=
#GOOGLE_SEARCH_ENGINE_ID=

```

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

### Environment Variable Configuration Example

Source: https://github.com/simpleyyt/ai-manus/blob/main/backend/README.md

Create a .env file and populate it with these environment variables to configure the service. This includes settings for model providers, LLM parameters, search APIs, sandbox environments, databases, and logging.

```bash
# Model provider configuration
API_KEY=your_api_key_here                # API key for model providers
API_BASE=https://api.openai.com/v1       # Base URL for model API (optional for some providers)

# Model configuration
MODEL_NAME=gpt-4o                        # Model name to use
MODEL_PROVIDER=openai                    # Model provider for LangChain
TEMPERATURE=0.7                          # Model temperature parameter
MAX_TOKENS=2000                          # Maximum output tokens per model request

# Google search configuration
GOOGLE_SEARCH_API_KEY=                   # Google Search API key for web search functionality (optional)
GOOGLE_SEARCH_ENGINE_ID=                 # Google custom search engine ID (optional)

# Sandbox configuration
SANDBOX_IMAGE=simpleyyt/manus-sandbox          # Sandbox environment Docker image
SANDBOX_NAME_PREFIX=sandbox              # Sandbox container name prefix
SANDBOX_TTL_MINUTES=30                   # Sandbox container time-to-live (minutes)
SANDBOX_NETWORK=manus-network            # Docker network name for communication between sandbox containers

# Database configuration
MONGODB_URI=mongodb://localhost:27017    # MongoDB connection URL
MONGODB_DATABASE=manus                   # MongoDB database name
REDIS_HOST=localhost                     # Redis host
REDIS_PORT=6379                          # Redis port
REDIS_DB=0                               # Redis DB index

# Log configuration
LOG_LEVEL=INFO                           # Log level, options: DEBUG, INFO, WARNING, ERROR, CRITICAL
```

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

### Start AI Manus Services with Docker Compose

Source: https://github.com/simpleyyt/ai-manus/blob/main/README.md

Command to start all defined services in detached mode. Ensure the docker compose file is saved as 'docker-compose.yml' in the current directory.

```shell
docker compose up -d
```

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

### Backend Python Dependencies

Source: https://github.com/simpleyyt/ai-manus/blob/main/AGENTS.md

Install backend dependencies using uv. Ensure you are in the backend directory.

```bash
cd backend && uv sync
```

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

### Run Development Server

Source: https://github.com/simpleyyt/ai-manus/blob/main/backend/README.md

Start the FastAPI development server with hot-reloading enabled for rapid development. The service will be accessible at http://localhost:8000.

```bash
# Start the development server (with hot reload)
uv run uvicorn app.main:app --host 0.0.0.0 --port 8000 --reload
```

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

### SerpAPI Custom Search Configuration (GET)

Source: https://github.com/simpleyyt/ai-manus/blob/main/docs/en/configuration.md

Configure AI Manus to use SerpAPI as a custom search provider via GET requests. Requires a SerpAPI key and specifies the query parameter for authentication.

```env
SEARCH_PROVIDER=custom
SEARCH_API_URL=https://serpapi.com/search
SEARCH_API_KEY=your-serpapi-key
SEARCH_API_KEY_PARAM=api_key
SEARCH_API_METHOD=GET
SEARCH_RESULT_FIELD=organic_results
```

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

### Frontend TypeScript Dependencies

Source: https://github.com/simpleyyt/ai-manus/blob/main/AGENTS.md

Install frontend dependencies using npm. Ensure you are in the frontend directory.

```bash
cd frontend && npm install
```

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

### Run Backend Service Individually

Source: https://github.com/simpleyyt/ai-manus/blob/main/AGENTS.md

Starts the FastAPI backend service with hot-reloading outside of Docker. Requires MongoDB and Redis to be running separately. Ensure the API_KEY environment variable is set.

```bash
cd backend
uv sync
uv run uvicorn app.main:app --host 0.0.0.0 --port 8000 --reload
```

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

### Brave Search API Custom Search Configuration (GET)

Source: https://github.com/simpleyyt/ai-manus/blob/main/docs/en/configuration.md

Configure AI Manus to use Brave Search API as a custom search provider via GET requests. Requires a Brave API key and specifies the header for authentication.

```env
SEARCH_PROVIDER=custom
SEARCH_API_URL=https://api.search.brave.com/res/v1/web/search
SEARCH_API_KEY=your-brave-key
SEARCH_API_KEY_HEADER=X-Subscription-Token
SEARCH_API_KEY_HEADER_PREFIX=
SEARCH_API_METHOD=GET
SEARCH_RESULT_FIELD=web.results
SEARCH_SNIPPET_FIELD=description
```

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

### Run AI Manus in Debug Mode

Source: https://github.com/simpleyyt/ai-manus/blob/main/README.md

Starts all services using the development Docker Compose file, enabling automatic code reloading for faster development cycles. This command is equivalent to 'docker compose -f docker-compose-development.yaml up'.

```shell
./dev.sh up
```

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

### Wait for Process Response

Source: https://github.com/simpleyyt/ai-manus/blob/main/sandbox/README.md

Example JSON response indicating that a process has completed. It includes the session ID, return code, and final status.

```json
{
  "success": true,
  "message": "Process completed, return code: 0",
  "data": {
    "session_id": "session_id",
    "returncode": 0,
    "status": "completed"
  }
}
```

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

### Clone AI Manus Repository

Source: https://github.com/simpleyyt/ai-manus/blob/main/README.md

Clones the AI Manus project from GitHub and navigates into the project directory. This is the first step for local development setup.

```shell
git clone https://github.com/simpleyyt/ai-manus.git
cd ai-manus
```

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

### Docker Compose Configuration for MCP

Source: https://github.com/simpleyyt/ai-manus/blob/main/docs/mcp.md

Shows how to configure the MCP configuration file path within a Docker Compose setup for the AI Manus backend service.

```yaml
...
services:
  backend:
    image: simpleyyt/manus-backend
    volumes:
      - ./mcp.json:/etc/mcp.json  # 挂载 MCP 配置文件
      - ...
    environment:
      # MCP 配置文件路径
      - MCP_CONFIG_PATH=/etc/mcp.json
...
```

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

### Execute Shell Command Response

Source: https://github.com/simpleyyt/ai-manus/blob/main/sandbox/README.md

Example JSON response after successfully executing a shell command. It includes the session ID, command executed, and its status.

```json
{
  "success": true,
  "message": "Command executed",
  "data": {
    "session_id": "session_id",
    "command": "ls -la",
    "status": "running"
  }
}
```

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

### Execute Shell Command Request

Source: https://github.com/simpleyyt/ai-manus/blob/main/sandbox/README.md

Example JSON payload for executing a shell command. The 'id' is optional for creating a new session, and 'exec_dir' specifies the working directory.

```json
{
  "id": "session_id",  /* Optional, automatically created if not provided */
  "exec_dir": "/path/to/dir",  /* Optional, command execution working directory (must use absolute path) */
  "command": "ls -la"  /* Command to execute */
}
```

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

### Wait for Process Request

Source: https://github.com/simpleyyt/ai-manus/blob/main/sandbox/README.md

Example JSON payload for waiting for a process in a shell session to complete. An optional 'seconds' parameter can limit the wait time.

```json
{
  "id": "session_id",  /* Target session ID */
  "seconds": 10  /* Optional, wait time (seconds) */
}
```

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

### Get Session Response

Source: https://github.com/simpleyyt/ai-manus/blob/main/backend/README.md

Response format for retrieving session information, including conversation history.

```json
{
  "code": 0,
  "msg": "success",
  "data": {
    "session_id": "string",
    "title": "string",
    "events": []
  }
}
```

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

### View Shell Session Content Request

Source: https://github.com/simpleyyt/ai-manus/blob/main/sandbox/README.md

Example JSON payload for viewing the content of a specific shell session. Requires the target session ID.

```json
{
  "id": "session_id"  /* Target session ID */
}
```

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

### Build Production Version

Source: https://github.com/simpleyyt/ai-manus/blob/main/frontend/README.md

Compile the project for production deployment.

```bash
npm run build
```

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

### Build Docker Image

Source: https://github.com/simpleyyt/ai-manus/blob/main/frontend/README.md

Create a Docker image for the AI chatbot Vue application.

```bash
docker build -t ai-chatbot-vue .
```

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

### GET /api/v1/supervisor/timeout/status

Source: https://github.com/simpleyyt/ai-manus/blob/main/sandbox/README.md

Retrieves the current status of the timeout function.

```APIDOC
## GET /api/v1/supervisor/timeout/status

### Description
Get the status of the timeout function.

### Method
GET

### Endpoint
/api/v1/supervisor/timeout/status

### Response
#### Success Response (200)
- **success** (boolean) - Indicates if the operation was successful.
- **message** (string) - A message describing the current timeout status.
- **data** (object) - Contains additional data about the timeout status.
  - **active** (boolean) - Indicates if the timeout function is currently active.
  - **timeout_timestamp** (string) - The timestamp when the timeout will occur if active.
  - **remaining_seconds** (integer) - The remaining time in seconds until the timeout occurs.

#### Response Example
```json
{
  "success": true,
  "message": "Remaining time: 45 minutes",
  "data": {
    "active": true,
    "timeout_timestamp": "2023-07-01T14:04:56",
    "remaining_seconds": 2700
  }
}
```
```

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

### Get Process Status

Source: https://github.com/simpleyyt/ai-manus/blob/main/sandbox/README.md

Retrieves the status of all service processes managed by the supervisor.

```APIDOC
## GET /api/v1/supervisor/status

### Description
Get the status of all service processes.

### Method
GET

### Endpoint
/api/v1/supervisor/status

### Response
#### Success Response (200)
- **success** (boolean) - Indicates if the operation was successful
- **message** (string) - A message describing the result
- **data** (array) - An array of service process statuses
  - **name** (string) - The name of the service
  - **status** (string) - The current status of the service (e.g., RUNNING)
  - **description** (string) - Additional details about the process (e.g., PID, uptime)

#### Response Example
```json
{
  "success": true,
  "message": "Services status retrieved successfully",
  "data": [
    {
      "name": "chrome",
      "status": "RUNNING",
      "description": "pid 123, uptime 10:30:45"
    }
  ]
}
```
```

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

### Development Environment Variables

Source: https://github.com/simpleyyt/ai-manus/blob/main/frontend/README.md

Configure the backend API URL for development. Create a `.env.development` file in the project root.

```env
VITE_API_URL=http://127.0.0.1:8000
```

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

### Build Docker Image for Production

Source: https://github.com/simpleyyt/ai-manus/blob/main/backend/README.md

Build the Docker image for the AI Manus × Claw backend service. This image can then be run as a container.

```bash
# Build Docker image
docker build -t manus-ai-agent .

```

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

### Get Session

Source: https://github.com/simpleyyt/ai-manus/blob/main/backend/README.md

Retrieves session information, including conversation history, using the session ID.

```APIDOC
## GET /api/v1/sessions/{session_id}

### Description
Get session information including conversation history.

### Method
GET

### Endpoint
/api/v1/sessions/{session_id}

### Parameters
#### Path Parameters
- **session_id** (string) - Required - The ID of the session to retrieve.

### Response
#### Success Response (200)
- **code** (integer) - 0 for success
- **msg** (string) - "success"
- **data** (object) - Contains session details
  - **session_id** (string) - The ID of the session.
  - **title** (string) - The title of the session.
  - **events** (array) - A list of conversation events.

### Response Example
```json
{
  "code": 0,
  "msg": "success",
  "data": {
    "session_id": "string",
    "title": "string",
    "events": []
  }
}
```
```

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

### VNC Connection

Source: https://github.com/simpleyyt/ai-manus/blob/main/backend/README.md

Establishes a VNC WebSocket connection to the session's sandbox environment.

```APIDOC
## WebSocket /api/v1/sessions/{session_id}/vnc

### Description
Establish a VNC WebSocket connection to the session's sandbox environment.

### Method
WebSocket

### Endpoint
/api/v1/sessions/{session_id}/vnc

### Parameters
#### Path Parameters
- **session_id** (string) - Required - The ID of the session.

### Protocol
WebSocket (binary mode)

### Subprotocol
`binary`
```

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

### View Backend Logs

Source: https://github.com/simpleyyt/ai-manus/blob/main/AGENTS.md

Follow the logs of the backend service for debugging purposes.

```bash
./dev.sh logs -f backend
```

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

### Run Docker Container

Source: https://github.com/simpleyyt/ai-manus/blob/main/backend/README.md

Run the AI Manus × Claw backend service as a Docker container. Ensure to map the Docker socket and use an .env file for configuration.

```bash
# Run container
docker run -p 8000:8000 --env-file .env -v /var/run/docker.sock:/var/run/docker.sock manus-ai-agent
```

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

### Get Process Status

Source: https://github.com/simpleyyt/ai-manus/blob/main/sandbox/README.md

Retrieves the status of all service processes managed by the supervisor. The response includes the name, status, and uptime for each service.

```json
{
  "success": true,
  "message": "Services status retrieved successfully",
  "data": [
    {
      "name": "chrome",
      "status": "RUNNING",
      "description": "pid 123, uptime 10:30:45"
    }
  ]
}
```

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

### Build Docker Image

Source: https://github.com/simpleyyt/ai-manus/blob/main/sandbox/README.md

Builds the Docker image for the manus-sandbox service. This command should be run from the root directory of the project.

```bash
docker build -t manus-sandbox .
```

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

### Basic AI Manus Environment Configuration

Source: https://github.com/simpleyyt/ai-manus/blob/main/README.md

Minimum required environment variables for AI Manus. At a minimum, set your OpenAI API key. Refer to .env.example or the documentation for all options.

```ini
API_KEY=sk-xxxx
API_BASE=https://api.openai.com/v1
MODEL_NAME=gpt-4o
```

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

### Build and Push AI Manus Docker Images

Source: https://github.com/simpleyyt/ai-manus/blob/main/README.md

Commands to build and push Docker images for AI Manus. Set the IMAGE_REGISTRY and IMAGE_TAG environment variables before running these commands.

```shell
export IMAGE_REGISTRY=your-registry-url
export IMAGE_TAG=latest

# Build images
./run build

# Push to the corresponding image repository
./run push
```

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

### Frontend TypeScript Type Check and Build

Source: https://github.com/simpleyyt/ai-manus/blob/main/AGENTS.md

Perform type checking and build the frontend application.

```bash
cd frontend && npm run type-check && npm run build
```

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

### Docker Compose Configuration for MCP

Source: https://github.com/simpleyyt/ai-manus/blob/main/docs/en/mcp.md

Shows how to configure the MCP service within a docker-compose.yml file, mounting the mcp.json configuration and setting the environment path.

```yaml
...
services:
  backend:
    image: simpleyyt/manus-backend
    volumes:
      - ./mcp.json:/etc/mcp.json  # Mount MCP configuration file
      - ...
    environment:
      # MCP configuration file path
      - MCP_CONFIG_PATH=/etc/mcp.json
...
```

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

### Connect to VNC

Source: https://github.com/simpleyyt/ai-manus/blob/main/sandbox/README.md

Connect to the VNC server to access the graphical interface of the container. This is typically done using a VNC client application.

```text
localhost:5900
```

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

### Get Service Timeout Status

Source: https://github.com/simpleyyt/ai-manus/blob/main/sandbox/README.md

Retrieve the current status of the timeout function, including whether it is active and the remaining time. The response provides the remaining time in seconds and the timestamp for the timeout.

```json
{
  "success": true,
  "message": "Remaining time: 45 minutes",
  "data": {
    "active": true,
    "timeout_timestamp": "2023-07-01T14:04:56",
    "remaining_seconds": 2700
  }
}
```

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

### Run Docker Container

Source: https://github.com/simpleyyt/ai-manus/blob/main/frontend/README.md

Run the Docker container and map the host port 8080 to the container port 80. Access the application via http://localhost:8080.

```bash
docker run -d -p 8080:80 ai-chatbot-vue
```

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

### Serper.dev Custom Search Configuration (POST)

Source: https://github.com/simpleyyt/ai-manus/blob/main/docs/en/configuration.md

Configure AI Manus to use Serper.dev as a custom search provider via POST requests. Requires a Serper.dev API key and specifies the header for authentication.

```env
SEARCH_PROVIDER=custom
SEARCH_API_URL=https://google.serper.dev/search
SEARCH_API_KEY=your-serper-key
SEARCH_API_KEY_HEADER=X-API-KEY
SEARCH_API_KEY_HEADER_PREFIX=
SEARCH_RESULT_FIELD=organic
```

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

### View File Content Response

Source: https://github.com/simpleyyt/ai-manus/blob/main/backend/README.md

Response format for viewing file content from the sandbox environment.

```json
{
  "code": 0,
  "msg": "success",
  "data": {
    "content": "file content",
    "file": "file path"
  }
}
```

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

### View File Content

Source: https://github.com/simpleyyt/ai-manus/blob/main/backend/README.md

Views the content of a file within the sandbox environment.

```APIDOC
## POST /api/v1/sessions/{session_id}/file

### Description
View file content in the sandbox environment.

### Method
POST

### Endpoint
/api/v1/sessions/{session_id}/file

### Parameters
#### Path Parameters
- **session_id** (string) - Required - The ID of the session.

### Request Body
- **file** (string) - Required - The path to the file.

### Request Example
```json
{
  "file": "file path"
}
```

### Response
#### Success Response (200)
- **code** (integer) - 0 for success
- **msg** (string) - "success"
- **data** (object) - Contains file content details
  - **content** (string) - The content of the file.
  - **file** (string) - The path to the file.

### Response Example
```json
{
  "code": 0,
  "msg": "success",
  "data": {
    "content": "file content",
    "file": "file path"
  }
}
```
```

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

### Run Docker Container

Source: https://github.com/simpleyyt/ai-manus/blob/main/sandbox/README.md

Runs the manus-sandbox Docker container, exposing necessary ports for the FastAPI service, Chrome debugging, and VNC.

```bash
docker run -p 8080:8080 -p 9222:9222 -p 5900:5900 -p 5901:5901 manus-sandbox
```

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

### MCP Server Configuration Structure

Source: https://github.com/simpleyyt/ai-manus/blob/main/docs/mcp.md

Defines the structure for configuring MCP servers, including command, arguments, transport, and environment variables.

```json
{
  "mcpServers": {
    "服务器名称": {
      "command": "命令",
      "args": ["参数列表"],
      "transport": "传输方式",
      "enabled": true/false,
      "description": "服务器描述",
      "env": {
        "环境变量名": "环境变量值"
      }
    }
  }
}
```

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

### Run Backend Integration Tests

Source: https://github.com/simpleyyt/ai-manus/blob/main/AGENTS.md

Executes backend integration tests using pytest. Requires the backend, MongoDB, and Redis to be running. Tests can be run for all files, specific files, or by marker.

```bash
# Ensure backend + MongoDB + Redis are running
./dev.sh up -d mongodb redis backend

cd backend
uv run pytest                               # all tests
uv run pytest tests/test_auth_routes.py     # specific file
uv run pytest -m file_api                   # by marker
```

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

### Docker Compose Backend Configuration

Source: https://github.com/simpleyyt/ai-manus/blob/main/docs/en/quick_start.md

This snippet shows how to configure the `backend` service in `docker-compose.yml` to use an environment file. This is an alternative to directly defining environment variables within the `docker-compose.yml` file.

```yaml
backend:
  image: simpleyyt/manus-backend
  # ...
  env_file:
    - .env
```

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

### List All Sessions

Source: https://github.com/simpleyyt/ai-manus/blob/main/backend/README.md

Retrieves a list of all available sessions.

```APIDOC
## GET /api/v1/sessions

### Description
Get list of all sessions.

### Method
GET

### Endpoint
/api/v1/sessions

### Response
#### Success Response (200)
- **code** (integer) - 0 for success
- **msg** (string) - "success"
- **data** (object) - Contains a list of sessions
  - **sessions** (array) - List of session objects
    - **session_id** (string) - The ID of the session.
    - **title** (string) - The title of the session.
    - **latest_message** (string) - The latest message in the session.
    - **latest_message_at** (integer) - Timestamp of the latest message.
    - **status** (string) - The current status of the session.
    - **unread_message_count** (integer) - Number of unread messages.

### Response Example
```json
{
  "code": 0,
  "msg": "success",
  "data": {
    "sessions": [
      {
        "session_id": "string",
        "title": "string",
        "latest_message": "string",
        "latest_message_at": 1234567890,
        "status": "string",
        "unread_message_count": 0
      }
    ]
  }
}
```
```

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

### Frontend Type Checking and Build

Source: https://github.com/simpleyyt/ai-manus/blob/main/AGENTS.md

Performs type checking on the frontend codebase using vue-tsc and creates a production build. These commands help catch TypeScript and template errors before deployment.

```bash
cd frontend
npm run type-check    # vue-tsc type checking
npm run build         # production build (catches TS + template errors)
```

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

### Write Input to Shell Process

Source: https://github.com/simpleyyt/ai-manus/blob/main/sandbox/README.md

Writes input to a specified process session and optionally simulates pressing Enter.

```APIDOC
## POST /api/v1/shell/write

### Description
Write input to the process in the specified session.

### Method
POST

### Endpoint
/api/v1/shell/write

### Request Body
- **id** (string) - Required - Target session ID
- **input** (string) - Required - Content to write
- **press_enter** (boolean) - Required - Whether to simulate pressing Enter after input

### Request Example
```json
{
  "id": "session_id",
  "input": "example input",
  "press_enter": true
}
```

### Response
#### Success Response (200)
- **success** (boolean) - Indicates if the operation was successful
- **message** (string) - A message describing the result
- **data** (object) - Contains details about the operation
  - **session_id** (string) - The ID of the session
  - **input** (string) - The input that was written

#### Response Example
```json
{
  "success": true,
  "message": "Input written",
  "data": {
    "session_id": "session_id",
    "input": "example input"
  }
}
```
```

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

### Run Sandbox Tests

Source: https://github.com/simpleyyt/ai-manus/blob/main/AGENTS.md

Executes tests for the sandbox service. Requires the sandbox container to be running. Uses pytest for testing.

```bash
./dev.sh up -d sandbox
cd sandbox
uv run pytest
```

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

### Write Input to Shell Process

Source: https://github.com/simpleyyt/ai-manus/blob/main/sandbox/README.md

Use this endpoint to send input to a running shell process in a specified session. You can also control whether to simulate pressing Enter after the input.

```json
{
  "id": "session_id",  /* Target session ID */
  "input": "example input",  /* Content to write */
  "press_enter": true  /* Whether to simulate pressing Enter after input */
}
```

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

### MCP Configuration File Structure

Source: https://github.com/simpleyyt/ai-manus/blob/main/docs/en/mcp.md

Defines the structure of the mcp.json file for configuring MCP servers, including command, arguments, transport, and environment variables.

```json
{
  "mcpServers": {
    "server_name": {
      "command": "command",
      "args": ["argument_list"],
      "transport": "transport_method",
      "enabled": true/false,
      "description": "server_description",
      "env": {
        "environment_variable_name": "environment_variable_value"
      }
    }
  }
}
```

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

### View Shell Session Content

Source: https://github.com/simpleyyt/ai-manus/blob/main/backend/README.md

Views the output of a shell session within the sandbox environment.

```APIDOC
## POST /api/v1/sessions/{session_id}/shell

### Description
View shell session output in the sandbox environment.

### Method
POST

### Endpoint
/api/v1/sessions/{session_id}/shell

### Parameters
#### Path Parameters
- **session_id** (string) - Required - The ID of the session.

### Request Body
- **session_id** (string) - Required - The ID of the shell session.

### Request Example
```json
{
  "session_id": "shell session ID"
}
```

### Response
#### Success Response (200)
- **code** (integer) - 0 for success
- **msg** (string) - "success"
- **data** (object) - Contains shell output details
  - **output** (string) - The shell output content.
  - **session_id** (string) - The ID of the shell session.
  - **console** (array) - List of console interactions
    - **ps1** (string) - The prompt string.
    - **command** (string) - The executed command.
    - **output** (string) - The output of the command.

### Response Example
```json
{
  "code": 0,
  "msg": "success",
  "data": {
    "output": "shell output content",
    "session_id": "shell session ID",
    "console": [
      {
        "ps1": "prompt string",
        "command": "executed command",
        "output": "command output"
      }
    ]
  }
}
```
```

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

### View File Content Request Body

Source: https://github.com/simpleyyt/ai-manus/blob/main/backend/README.md

Request body for viewing file content in the sandbox environment.

```json
{
  "file": "file path"
}
```

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

### Backend Pytest Execution

Source: https://github.com/simpleyyt/ai-manus/blob/main/AGENTS.md

Run backend tests using pytest. This requires a running backend, MongoDB, and Redis.

```bash
cd backend && uv run pytest
```

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

### Find Files by Pattern

Source: https://github.com/simpleyyt/ai-manus/blob/main/sandbox/README.md

Finds files within a specified directory based on a glob pattern. Returns a list of matching file paths.

```json
{
  "path": "/path/to/dir",  /* Directory path to search */
  "glob": "*.txt"  /* Filename pattern (glob syntax) */
}
```

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

### Create Session Response

Source: https://github.com/simpleyyt/ai-manus/blob/main/backend/README.md

Response format for successfully creating a new conversation session.

```json
{
  "code": 0,
  "msg": "success",
  "data": {
    "session_id": "string"
  }
}
```

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

### Docker Compose Configuration for AI Manus

Source: https://github.com/simpleyyt/ai-manus/blob/main/README.md

This YAML file defines the services, networks, and volumes required to run the AI Manus application using Docker Compose. It includes configurations for frontend, backend, sandbox, claw, MongoDB, and Redis.

```yaml
services:
  frontend:
    image: simpleyyt/manus-frontend
    ports:
      - "5173:80"
    depends_on:
      - backend
    restart: unless-stopped
    networks:
      - manus-network
    environment:
      - BACKEND_URL=http://backend:8000

  backend:
    image: simpleyyt/manus-backend
    depends_on:
      - sandbox
      - claw
    restart: unless-stopped
    volumes:
      - /var/run/docker.sock:/var/run/docker.sock:ro
      #- ./mcp.json:/etc/mcp.json # Mount MCP servers directory
    networks:
      - manus-network
    environment:
      # OpenAI API base URL
      - API_BASE=https://api.openai.com/v1
      # OpenAI API key, replace with your own
      - API_KEY=sk-xxxx
      # LLM model name
      - MODEL_NAME=gpt-4o
      # LLM temperature parameter, controls randomness
      #- TEMPERATURE=0.7
      # Maximum tokens for LLM response
      #- MAX_TOKENS=2000
      # More configuration options: https://docs.ai-manus.com/#/configuration

  sandbox:
    image: simpleyyt/manus-sandbox
    command: /bin/sh -c "exit 0"  # prevent sandbox from starting, ensure image is pulled
    restart: "no"
    networks:
      - manus-network

  claw:
    image: simpleyyt/manus-claw
    entrypoint: /bin/sh -c "exit 0"  # prevent claw from starting, ensure image is pulled
    restart: "no"
    networks:
      - manus-network

  mongodb:
    image: mongo:7.0
    volumes:
      - mongodb_data:/data/db
    restart: unless-stopped
    #ports:
    #  - "27017:27017"
    networks:
      - manus-network

  redis:
    image: redis:7.0
    restart: unless-stopped
    networks:
      - manus-network

volumes:
  mongodb_data:
    name: manus-mongodb-data

networks:
  manus-network:
    name: manus-network
    driver: bridge

```

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

### POST /api/v1/supervisor/restart

Source: https://github.com/simpleyyt/ai-manus/blob/main/sandbox/README.md

Restarts all services managed by the supervisor.

```APIDOC
## POST /api/v1/supervisor/restart

### Description
Restart all services.

### Method
POST

### Endpoint
/api/v1/supervisor/restart

### Response
#### Success Response (200)
- **success** (boolean) - Indicates if the operation was successful.
- **message** (string) - A message describing the result of the operation.
- **data** (object) - Contains additional data about the operation.
  - **status** (string) - The status of the services after restart.

#### Response Example
```json
{
  "success": true,
  "message": "All services restarted",
  "data": {
    "status": "restarted"
  }
}
```
```

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

### Sandbox Pytest Execution

Source: https://github.com/simpleyyt/ai-manus/blob/main/AGENTS.md

Run sandbox tests using pytest. Ensure you are in the sandbox directory.

```bash
cd sandbox && uv run pytest
```

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

### Clean Up and Rebuild AI Manus Development Environment

Source: https://github.com/simpleyyt/ai-manus/blob/main/README.md

Commands to clean up all related Docker resources and rebuild images for the development environment. Use this when dependencies change.

```shell
# Clean up all related resources
./dev.sh down -v

# Rebuild images
./dev.sh build

# Run in debug mode
./dev.sh up
```

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

### Mockserver API Test

Source: https://github.com/simpleyyt/ai-manus/blob/main/AGENTS.md

Verify the mockserver API by sending a POST request to the chat completions endpoint.

```bash
curl -X POST http://localhost:8090/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{"model":"mock","messages":[{"role":"user","content":"hi"}]}'
```

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

### List All Sessions Response

Source: https://github.com/simpleyyt/ai-manus/blob/main/backend/README.md

Response format for listing all available conversation sessions.

```json
{
  "code": 0,
  "msg": "success",
  "data": {
    "sessions": [
      {
        "session_id": "string",
        "title": "string",
        "latest_message": "string",
        "latest_message_at": 1234567890,
        "status": "string",
        "unread_message_count": 0
      }
    ]
  }
}
```

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

### Execute Shell Command

Source: https://github.com/simpleyyt/ai-manus/blob/main/sandbox/README.md

Executes a shell command within a specified or newly created session. It allows for defining the execution directory and returns the session details upon successful initiation.

```APIDOC
## POST /api/v1/shell/exec

### Description
Execute a command in the specified shell session.

### Method
POST

### Endpoint
/api/v1/shell/exec

### Request Body
- **id** (string) - Optional - session_id, automatically created if not provided
- **exec_dir** (string) - Optional - command execution working directory (must use absolute path)
- **command** (string) - Required - Command to execute

### Response
#### Success Response (200)
- **success** (boolean) - Indicates if the operation was successful.
- **message** (string) - A message describing the result of the operation.
- **data** (object) - Contains details about the executed command.
  - **session_id** (string) - The ID of the shell session.
  - **command** (string) - The command that was executed.
  - **status** (string) - The current status of the command execution (e.g., "running").

### Request Example
```json
{
  "id": "session_id",
  "exec_dir": "/path/to/dir",
  "command": "ls -la"
}
```

### Response Example
```json
{
  "success": true,
  "message": "Command executed",
  "data": {
    "session_id": "session_id",
    "command": "ls -la",
    "status": "running"
  }
}
```
```

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

### Search File Content

Source: https://github.com/simpleyyt/ai-manus/blob/main/sandbox/README.md

Searches file content using regular expressions, returning matching lines and optional sudo permissions.

```APIDOC
## POST /api/v1/file/search

### Description
Search file content using regular expressions.

### Method
POST

### Endpoint
/api/v1/file/search

### Request Body
- **file** (string) - Required - Absolute file path
- **regex** (string) - Required - Regular expression pattern
- **sudo** (boolean) - Optional - Whether to use sudo permissions

### Request Example
```json
{
  "file": "/path/to/file",
  "regex": "search pattern",
  "sudo": false
}
```

### Response
#### Success Response (200)
- **success** (boolean) - Indicates if the operation was successful
- **message** (string) - A message describing the result, including the number of matches found
- **data** (object) - Contains details about the search results
  - **file** (string) - The path to the file
  - **matches** (array) - An array of matching occurrences
    - **line_number** (integer) - The line number where the match occurred
    - **line** (string) - The content of the line containing the match
    - **match** (string) - The specific matched content

#### Response Example
```json
{
  "success": true,
  "message": "Search completed, found 3 matches",
  "data": {
    "file": "/path/to/file",
    "matches": [
      {
        "line_number": 10,
        "line": "Matching line content",
        "match": "Matching content"
      }
    ]
  }
}
```
```

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

### Search File Content

Source: https://github.com/simpleyyt/ai-manus/blob/main/sandbox/README.md

Searches for content within a file using regular expressions. Returns matching lines and their details. Sudo permissions can be optionally used.

```json
{
  "file": "/path/to/file",  /* Absolute file path */
  "regex": "search pattern",  /* Regular expression pattern */
  "sudo": false  /* Optional, whether to use sudo permissions */
}
```

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

### Create Session

Source: https://github.com/simpleyyt/ai-manus/blob/main/backend/README.md

Creates a new conversation session. This endpoint does not require a request body.

```APIDOC
## PUT /api/v1/sessions

### Description
Create a new conversation session.

### Method
PUT

### Endpoint
/api/v1/sessions

### Request Body
None

### Response
#### Success Response (200)
- **code** (integer) - 0 for success
- **msg** (string) - "success"
- **data** (object) - Contains session details
  - **session_id** (string) - The ID of the newly created session

### Response Example
```json
{
  "code": 0,
  "msg": "success",
  "data": {
    "session_id": "string"
  }
}
```
```

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

### Wait for Process

Source: https://github.com/simpleyyt/ai-manus/blob/main/sandbox/README.md

Waits for a process within a specified shell session to complete. An optional timeout can be provided to limit the waiting period.

```APIDOC
## POST /api/v1/shell/wait

### Description
Wait for the process in the specified session to complete.

### Method
POST

### Endpoint
/api/v1/shell/wait

### Request Body
- **id** (string) - Required - Target session ID
- **seconds** (integer) - Optional - wait time (seconds)

### Response
#### Success Response (200)
- **success** (boolean) - Indicates if the operation was successful.
- **message** (string) - A message describing the result of the operation, including the return code.
- **data** (object) - Contains details about the completed process.
  - **session_id** (string) - The ID of the shell session.
  - **returncode** (integer) - The exit code of the completed process.
  - **status** (string) - The final status of the process (e.g., "completed").

### Request Example
```json
{
  "id": "session_id",
  "seconds": 10
}
```

### Response Example
```json
{
  "success": true,
  "message": "Process completed, return code: 0",
  "data": {
    "session_id": "session_id",
    "returncode": 0,
    "status": "completed"
  }
}
```
```

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

### View Shell Session Content Request Body

Source: https://github.com/simpleyyt/ai-manus/blob/main/backend/README.md

Request body for viewing shell session output in the sandbox environment.

```json
{
  "session_id": "shell session ID"
}
```

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

### Write File Content

Source: https://github.com/simpleyyt/ai-manus/blob/main/sandbox/README.md

Writes content to a specified file. Supports options for append mode, adding leading/trailing newlines, and using sudo permissions.

```json
{
  "file": "/path/to/file",  /* Absolute file path */
  "content": "File content",  /* Content to write */
  "append": false,  /* Optional, whether to use append mode */
  "leading_newline": false,  /* Optional, whether to add a newline before the content */
  "trailing_newline": false,  /* Optional, whether to add a newline after the content */
  "sudo": false  /* Optional, whether to write with sudo permissions */
}
```