### Install Ornold MCP with npm

Source: https://github.com/ornold-ai/ornold-mcp/blob/main/CLAUDE.md

Use this command to install the Ornold MCP npm package. A user token from mcp.ornold.com is required.

```bash
npx mcp-ornold-browser --token USER_TOKEN
```

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

### Multi-Browser Configuration Example

Source: https://context7.com/ornold-ai/ornold-mcp/llms.txt

Configure multiple antidetect browsers simultaneously for parallel automation across different browser platforms. Includes ports and tokens for Linken Sphere, Dolphin, and Vision.

```bash
npx mcp-ornold-browser \
  --token orn_sk_your_token \
  --mode both \
  --linken-port 40080 \
  --dolphin-port 3001 \
  --dolphin-token eyJhbGciOiJIUzI1NiJ9... \
  --vision-token vx_your_token \
  --wadex-port 8080
```

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

### Start Browser Recording

Source: https://context7.com/ornold-ai/ornold-mcp/llms.txt

Begin recording a browser automation flow. The recorded flow can be saved and replayed across multiple browser profiles.

```javascript
// Start recording a new flow
await mcp.callTool("browser_start_recording", {
  "name": "login_flow"
});
```

```javascript
// Expected response:
// {
//   "recording": true,
//   "flow_id": "flow_abc123",
//   "name": "login_flow"
// }
```

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

### Full Linken Sphere Setup with Vision Mode

Source: https://context7.com/ornold-ai/ornold-mcp/llms.txt

Complete configuration for antidetect browser automation with Linken Sphere using vision mode for maximum stealth. Requires a production token.

```bash
# Start MCP server with full Linken Sphere integration
npx mcp-ornold-browser \
  --token orn_sk_your_production_token \
  --mode vision \
  --linken-port 40080
```

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

### Full Linken Sphere Setup with Vision Mode

Source: https://context7.com/ornold-ai/ornold-mcp/llms.txt

Complete configuration for antidetect browser automation with Linken Sphere using vision mode for maximum stealth.

```APIDOC
## Full Linken Sphere Setup with Vision Mode

Complete configuration for antidetect browser automation with Linken Sphere using vision mode for maximum stealth.

### Command Example
```bash
npx mcp-ornold-browser \
  --token orn_sk_your_production_token \
  --mode vision \
  --linken-port 40080
```

### Claude Code Configuration Example
```json
// Claude Code config with all options
// Run: claude mcp add --transport stdio ornold-browser -- npx mcp-ornold-browser --token orn_sk_token --mode vision --linken-port 40080
```
```

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

### Install Ornold MCP with Claude Code

Source: https://github.com/ornold-ai/ornold-mcp/blob/main/README.md

Install the Ornold MCP browser automation package using Claude Code. Ensure you have your token from mcp.ornold.com.

```bash
claude mcp add --transport stdio ornold-browser -- npx mcp-ornold-browser --token YOUR_TOKEN
```

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

### Install AI Skill for Claude Code

Source: https://github.com/ornold-ai/ornold-mcp/blob/main/README.md

After installing Ornold MCP with Claude Code, install the AI skill to enable your agent to effectively use the tools.

```bash
npx mcp-ornold-browser --install-skill
```

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

### Install Ornold MCP for Claude Code

Source: https://github.com/ornold-ai/ornold-mcp/blob/main/AGENTS.md

Use these commands to install Ornold MCP when operating as Claude Code. Ensure you have your Ornold token ready.

```bash
claude mcp add --transport stdio ornold-browser -- npx mcp-ornold-browser --token TOKEN
```

```bash
npx mcp-ornold-browser --install-skill
```

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

### Install Ornold MCP in VS Code via CLI

Source: https://context7.com/ornold-ai/ornold-mcp/llms.txt

Add Ornold MCP to VS Code using the command line interface. This command registers the Ornold browser with the specified token.

```bash
# CLI installation
code --add-mcp '{"name":"ornold-browser","command":"npx","args":["mcp-ornold-browser","--token","orn_sk_your_token_here"]}'
```

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

### Add Ornold MCP to Claude Code

Source: https://context7.com/ornold-ai/ornold-mcp/llms.txt

Integrate Ornold MCP into Claude Code using stdio transport and token authentication. Optional AI skill installation enhances agent capabilities.

```bash
# Add the MCP server to Claude Code
claude mcp add --transport stdio ornold-browser -- npx mcp-ornold-browser --token orn_sk_your_token_here

# Install the AI skill (teaches the agent to use tools effectively)
npx mcp-ornold-browser --install-skill
```

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

### browser_start_recording

Source: https://context7.com/ornold-ai/ornold-mcp/llms.txt

Begin recording a browser automation flow that can be saved and replayed across multiple browser profiles.

```APIDOC
## browser_start_recording

Begin recording a browser automation flow that can be saved and replayed across multiple browser profiles.

### Method
```javascript
await mcp.callTool("browser_start_recording", {
  "name": "login_flow"
})
```

### Parameters
#### Request Body
- **name** (string) - Required - The name for the automation flow.

### Response Example (Success)
```json
{
  "recording": true,
  "flow_id": "flow_abc123",
  "name": "login_flow"
}
```
```

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

### Connect to Wadex Browser

Source: https://context7.com/ornold-ai/ornold-mcp/llms.txt

Connect to the Wadex antidetect browser by specifying its local API port for session management.

```bash
npx mcp-ornold-browser --token orn_sk_token --wadex-port 8080
```

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

### Wadex Configuration

Source: https://context7.com/ornold-ai/ornold-mcp/llms.txt

Connect to Wadex antidetect browser by specifying the local API port for session management.

```APIDOC
## Wadex Configuration

Connect to Wadex antidetect browser by specifying the local API port for session management.

### Command
```bash
npx mcp-ornold-browser --token orn_sk_token --wadex-port 8080
```
```

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

### Multi-Browser Configuration

Source: https://context7.com/ornold-ai/ornold-mcp/llms.txt

Configure multiple antidetect browsers simultaneously for parallel automation across different browser platforms.

```APIDOC
## Multi-Browser Configuration

Configure multiple antidetect browsers simultaneously for parallel automation across different browser platforms.

### Command Example
```bash
npx mcp-ornold-browser \
  --token orn_sk_your_token \
  --mode both \
  --linken-port 40080 \
  --dolphin-port 3001 \
  --dolphin-token eyJhbGciOiJIUzI1NiJ9... \
  --vision-token vx_your_token \
  --wadex-port 8080
```
```

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

### Connect to Vision Browser with Ornold MCP

Source: https://context7.com/ornold-ai/ornold-mcp/llms.txt

Integrate Ornold MCP with Vision Browser using either the X-Token for cloud management or a local port for direct connection, or both.

```bash
# Cloud connection with token
npx mcp-ornold-browser --token orn_sk_token --vision-token vx_your_vision_token

# Local connection with port
npx mcp-ornold-browser --token orn_sk_token --vision-port 8080

# Both local and cloud
npx mcp-ornold-browser --token orn_sk_token --vision-token vx_token --vision-port 8080
```

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

### browser_status

Source: https://context7.com/ornold-ai/ornold-mcp/llms.txt

Verify the MCP server is running and check connection status to configured antidetect browsers.

```APIDOC
## browser_status

Verify the MCP server is running and check connection status to configured antidetect browsers.

### Method
```javascript
await mcp.callTool("browser_status", {{}})
```

### Response Example (Success)
```json
{
  "status": "connected",
  "mode": "vision",
  "browsers": {
    "linken_sphere": { "connected": true, "profiles": 12 },
    "dolphin_anty": { "connected": false }
  }
}
```
```

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

### Configure Ornold MCP in Cursor

Source: https://context7.com/ornold-ai/ornold-mcp/llms.txt

Set up Ornold MCP in Cursor by writing the server configuration to the Cursor config file. Ensure the token is correctly specified.

```json
// ~/.cursor/mcp.json
{
  "mcpServers": {
    "ornold-browser": {
      "command": "npx",
      "args": ["mcp-ornold-browser", "--token", "orn_sk_your_token_here"]
    }
  }
}
```

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

### Configure Ornold MCP for Cursor

Source: https://github.com/ornold-ai/ornold-mcp/blob/main/README.md

Add Ornold MCP server configuration to Cursor's mcp.json file. Replace YOUR_TOKEN with your actual Ornold token.

```json
{
  "mcpServers": {
    "ornold-browser": {
      "command": "npx",
      "args": ["mcp-ornold-browser", "--token", "YOUR_TOKEN"]
    }
  }
}
```

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

### Configure Ornold MCP for Cursor

Source: https://github.com/ornold-ai/ornold-mcp/blob/main/AGENTS.md

For Cursor users, configure Ornold MCP by writing the provided JSON to `~/.cursor/mcp.json`. Replace 'TOKEN' with the actual Ornold token.

```json
{
  "mcpServers": {
    "ornold-browser": {
      "command": "npx",
      "args": ["mcp-ornold-browser", "--token", "TOKEN"]
    }
  }
}
```

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

### Configure Ornold MCP for VS Code (CLI)

Source: https://github.com/ornold-ai/ornold-mcp/blob/main/README.md

Add Ornold MCP server configuration to VS Code using the command line. This registers the 'ornold-browser' MCP server.

```bash
code --add-mcp '{"name":"ornold-browser","command":"npx","args":["mcp-ornold-browser","--token","YOUR_TOKEN"]}'
```

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

### Configure Ornold MCP in VS Code

Source: https://context7.com/ornold-ai/ornold-mcp/llms.txt

Set up Ornold MCP in VS Code by creating a configuration file in your workspace. This specifies the server type, command, and arguments.

```json
// .vscode/mcp.json
{
  "servers": {
    "ornold-browser": {
      "type": "stdio",
      "command": "npx",
      "args": ["mcp-ornold-browser", "--token", "orn_sk_your_token_here"]
    }
  }
}
```

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

### Configure Ornold MCP for Codex

Source: https://github.com/ornold-ai/ornold-mcp/blob/main/AGENTS.md

For Codex users, configure Ornold MCP by writing the provided TOML to `~/.codex/config.toml`. Replace 'TOKEN' with the user's Ornold token.

```toml
[mcp_servers.ornold-browser]
command = "npx"
args = ["-y", "mcp-ornold-browser", "--token", "TOKEN"]
```

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

### Check Browser Status

Source: https://context7.com/ornold-ai/ornold-mcp/llms.txt

Verify the MCP server is running and check the connection status to configured antidetect browsers. Expected response includes connection status and profile counts.

```javascript
await mcp.callTool("browser_status", {});
```

```javascript
// Expected response:
// {
//   "status": "connected",
//   "mode": "vision",
//   "browsers": {
//     "linken_sphere": { "connected": true, "profiles": 12 },
//     "dolphin_anty": { "connected": false }
//   }
// }
```

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

### Connect to Linken Sphere with Ornold MCP

Source: https://context7.com/ornold-ai/ornold-mcp/llms.txt

Integrate Ornold MCP with Linken Sphere antidetect browser by specifying the local API port. This enables full session management and proxy configuration.

```bash
npx mcp-ornold-browser --token orn_sk_token --linken-port 40080
```

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

### Add Ornold MCP to Windsurf

Source: https://context7.com/ornold-ai/ornold-mcp/llms.txt

Integrate Ornold MCP into Windsurf via Settings > Tools > Add MCP Server. Provide the server configuration including command and arguments.

```json
{
  "mcpServers": {
    "ornold-browser": {
      "command": "npx",
      "args": ["mcp-ornold-browser", "--token", "orn_sk_your_token_here"]
    }
  }
}
```

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

### Connect to Dolphin Anty with Ornold MCP

Source: https://context7.com/ornold-ai/ornold-mcp/llms.txt

Connect Ornold MCP to Dolphin Anty using local port and optionally a cloud API token for comprehensive profile management.

```bash
# Local API only
npx mcp-ornold-browser --token orn_sk_token --dolphin-port 3001

# Local + Cloud API
npx mcp-ornold-browser --token orn_sk_token --dolphin-port 3001 --dolphin-token eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9...
```

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

### Configure Ornold MCP for VS Code / Copilot

Source: https://github.com/ornold-ai/ornold-mcp/blob/main/AGENTS.md

Use this command to add Ornold MCP configuration for VS Code or Copilot. Replace 'TOKEN' with the user's Ornold token.

```bash
code --add-mcp '{"name":"ornold-browser","command":"npx","args":["mcp-ornold-browser","--token","TOKEN"]}'
```

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

### Configure Ornold MCP for VS Code (.vscode/mcp.json)

Source: https://github.com/ornold-ai/ornold-mcp/blob/main/README.md

Add Ornold MCP server configuration to your VS Code project by creating or editing the .vscode/mcp.json file. Specify the command and arguments for the 'ornold-browser' server.

```json
{
  "servers": {
    "ornold-browser": {
      "type": "stdio",
      "command": "npx",
      "args": ["mcp-ornold-browser", "--token", "YOUR_TOKEN"]
    }
  }
}
```

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

### Ornold MCP Interaction Mode Selection

Source: https://context7.com/ornold-ai/ornold-mcp/llms.txt

Select the interaction mode for browser automation. 'dom' is default, 'vision' uses AI for stealth, and 'both' allows per-task selection.

```bash
# DOM mode (default) - fast, uses DOM snapshots, good for simple pages
npx mcp-ornold-browser --token orn_sk_token --mode dom

# Vision mode - uses screenshots + AI element detection, required for antidetect stealth
npx mcp-ornold-browser --token orn_sk_token --mode vision

# Both modes - choose per-task based on requirements
npx mcp-ornold-browser --token orn_sk_token --mode both
```

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

### Configure Ornold MCP for Codex

Source: https://github.com/ornold-ai/ornold-mcp/blob/main/README.md

Add Ornold MCP server configuration to Codex by editing the ~/.codex/config.toml file. This TOML configuration specifies the command and arguments for the 'ornold-browser' MCP server.

```toml
[mcp_servers.ornold-browser]
command = "npx"
args = ["-y", "mcp-ornold-browser", "--token", "YOUR_TOKEN"]
```

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

### Configure Ornold MCP in Cline

Source: https://context7.com/ornold-ai/ornold-mcp/llms.txt

Set up Ornold MCP in Cline through the sidebar MCP settings. Ensure the 'disabled' flag is set to false for the Ornold browser configuration.

```json
{
  "mcpServers": {
    "ornold-browser": {
      "command": "npx",
      "args": ["mcp-ornold-browser", "--token", "orn_sk_your_token_here"],
      "disabled": false
    }
  }
}
```

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

### browser_save_flow

Source: https://context7.com/ornold-ai/ornold-mcp/llms.txt

Save the recorded flow for later replay across multiple browser profiles with unique data per profile.

```APIDOC
## browser_save_flow

Save the recorded flow for later replay across multiple browser profiles with unique data per profile.

### Method
```javascript
await mcp.callTool("browser_save_flow", {
  "flow_id": "flow_abc123",
  "name": "login_automation"
})
```

### Parameters
#### Request Body
- **flow_id** (string) - Required - The ID of the flow to save.
- **name** (string) - Required - The name to assign to the saved flow.

### Response Example (Success)
```json
{
  "saved": true,
  "flow_id": "flow_abc123",
  "name": "login_automation",
  "actions": 15
}
```
```

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

### Configure Ornold MCP in Codex

Source: https://context7.com/ornold-ai/ornold-mcp/llms.txt

Add Ornold MCP to Codex CLI and Desktop by configuring the TOML file. This specifies the command and arguments for the Ornold browser.

```toml
# ~/.codex/config.toml
[mcp_servers.ornold-browser]
command = "npx"
args = ["-y", "mcp-ornold-browser", "--token", "orn_sk_your_token_here"]
```

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

### Save Recorded Browser Flow

Source: https://context7.com/ornold-ai/ornold-mcp/llms.txt

Save the currently recorded browser automation flow. The saved flow can be replayed with unique data per profile.

```javascript
// Save the recorded flow
await mcp.callTool("browser_save_flow", {
  "flow_id": "flow_abc123",
  "name": "login_automation"
});
```

```javascript
// Expected response:
// {
//   "saved": true,
//   "flow_id": "flow_abc123",
//   "name": "login_automation",
//   "actions": 15
// }
```

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

### Configure Ornold MCP for Cline

Source: https://github.com/ornold-ai/ornold-mcp/blob/main/README.md

Add Ornold MCP server configuration in Cline via the MCP section in the sidebar. This JSON snippet defines the 'ornold-browser' server with its command and arguments.

```json
{
  "mcpServers": {
    "ornold-browser": {
      "command": "npx",
      "args": ["mcp-ornold-browser", "--token", "YOUR_TOKEN"],
      "disabled": false
    }
  }
}
```

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

### browser_stop_recording

Source: https://context7.com/ornold-ai/ornold-mcp/llms.txt

Stop the current recording session and prepare the flow for saving.

```APIDOC
## browser_stop_recording

Stop the current recording session and prepare the flow for saving.

### Method
```javascript
await mcp.callTool("browser_stop_recording", {{}})
```

### Response Example (Success)
```json
{
  "recording": false,
  "flow_id": "flow_abc123",
  "actions_recorded": 15
}
```
```

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

### Configure Ornold MCP for Cline

Source: https://github.com/ornold-ai/ornold-mcp/blob/main/AGENTS.md

Cline users can add Ornold MCP by pasting this JSON into the MCP configuration in the Cline sidebar. Replace 'TOKEN' with the user's Ornold token.

```json
{
  "mcpServers": {
    "ornold-browser": {
      "command": "npx",
      "args": ["mcp-ornold-browser", "--token", "TOKEN"],
      "disabled": false
    }
  }
}
```

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

### Ornold MCP Interaction Modes

Source: https://github.com/ornold-ai/ornold-mcp/blob/main/AGENTS.md

Specify the interaction mode for Ornold MCP using the `--mode` flag. 'vision' mode is recommended for antidetect browser tasks.

```bash
--mode dom
```

```bash
--mode vision
```

```bash
--mode both
```

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

### Ornold MCP Token Authentication

Source: https://context7.com/ornold-ai/ornold-mcp/llms.txt

Use the '--token' flag with your API token to authenticate with the Ornold service. Tokens are obtained from the Ornold dashboard.

```bash
npx mcp-ornold-browser --token orn_sk_your_token_here
```

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

### Claude Code Config for Linken Sphere

Source: https://context7.com/ornold-ai/ornold-mcp/llms.txt

Configuration for Claude Code to run MCP with Linken Sphere in vision mode. This command adds the Ornold browser and specifies the connection details.

```bash
// Claude Code config with all options
// Run: claude mcp add --transport stdio ornold-browser -- npx mcp-ornold-browser --token orn_sk_token --mode vision --linken-port 40080
```

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

### Antidetect Browser Flags for Ornold MCP

Source: https://github.com/ornold-ai/ornold-mcp/blob/main/AGENTS.md

Optional flags to configure specific antidetect browsers with Ornold MCP. Choose the flag corresponding to the user's browser.

```bash
--linken-port 40080
```

```bash
--dolphin-port 3001 --dolphin-token eyJ...
```

```bash
--vision-token vx_...
```

```bash
--wadex-port 8080
```

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

### Stop Browser Recording

Source: https://context7.com/ornold-ai/ornold-mcp/llms.txt

Stop the current browser automation recording session and prepare the flow for saving. Records the number of actions taken.

```javascript
// Stop recording
await mcp.callTool("browser_stop_recording", {});
```

```javascript
// Expected response:
// {
//   "recording": false,
//   "flow_id": "flow_abc123",
//   "actions_recorded": 15
// }
```

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

### browser_solve_captcha

Source: https://context7.com/ornold-ai/ornold-mcp/llms.txt

Automatically detect and solve CAPTCHAs including reCAPTCHA v2/v3, hCaptcha, Cloudflare Turnstile, GeeTest, FunCaptcha, Amazon WAF, and PerimeterX.

```APIDOC
## browser_solve_captcha

Automatically detect and solve CAPTCHAs including reCAPTCHA v2/v3, hCaptcha, Cloudflare Turnstile, GeeTest, FunCaptcha, Amazon WAF, and PerimeterX.

### Method
```javascript
await mcp.callTool("browser_solve_captcha", {{}})
```

### Parameters
#### Request Body
- **(No parameters needed for automatic detection)**

### Response Example (Success)
```json
{
  "solved": true,
  "type": "recaptcha_v2",
  "duration_ms": 12500
}
```
```

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

### Solve CAPTCHA Automatically

Source: https://context7.com/ornold-ai/ornold-mcp/llms.txt

Automatically detect and solve various types of CAPTCHAs on the current page. No parameters are needed for automatic detection.

```javascript
// Solve CAPTCHA on current page
await mcp.callTool("browser_solve_captcha", {
  // Automatic detection - no parameters needed
});
```

```javascript
// Expected response:
// {
//   "solved": true,
//   "type": "recaptcha_v2",
//   "duration_ms": 12500
// }
```

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