Skip to main content

How to Integrate Octoparse with VS Code via MCP (Step-by-Step)

Updated today

Before You Begin

Make sure you have:

  • VS Code installed

  • GitHub Copilot access in VS Code

VS Code stores MCP server configuration in an mcp.json file, which can live either in your workspace at .vscode/mcp.json or in your user profile. VS Code also provides IntelliSense for this file.

Good to know
In VS Code, MCP support is part of the GitHub Copilot experience, not a separate standalone connector marketplace.

What This Integration Enables

After Octoparse is connected to VS Code via MCP, you can use natural language prompts in Copilot to:

  • Search Octoparse scraping templates

  • Create scraping tasks from templates

  • Start and stop cloud tasks

  • Check task execution status

  • Export data in CSV or JSON

  • Access Octoparse account-related tools

Step 1: Open VS Code and Set Up GitHub Copilot

Open VS Code and make sure GitHub Copilot Chat is available in your editor.

Microsoft’s getting-started documentation says access to GitHub Copilot is required for Copilot chat and agent experiences in VS Code.

Step 2: Open MCP Server Configuration

In VS Code, open the MCP server configuration.

VS Code documents MCP server setup through the MCP configuration file, mcp.json. This can be stored:

  • in your workspace: .vscode/mcp.json

  • or in your user profile configuration

If you want the Octoparse MCP server available only for one project, use the workspace file. If you want it available more broadly, use the user-level configuration.

Step 3: Add the Octoparse MCP Server

Add Octoparse to your mcp.json configuration.

Use the Octoparse MCP server URL:

{
  "mcpServers": {
    "octoparse": {
      "url": "https://mcp.octoparse.com"
    }
  }
}

VS Code’s official MCP configuration reference confirms that MCP servers are defined in mcp.json, and that the editor provides schema-aware support for this file.

Note
Depending on your VS Code version and MCP auth flow, your final configuration may vary slightly from the example above. Use the latest VS Code MCP schema and prompts shown in your editor as the source of truth.

Step 4: Save the Configuration

Save mcp.json.

If the file is valid, VS Code should recognize the MCP server configuration. If the configuration is malformed or placed in the wrong location, the server may not load correctly.

Step 5: Complete Authentication

If Octoparse requires authentication, VS Code will guide you through the authorization flow when the MCP server is used.

As of mid-2025, VS Code supports the full MCP specification, including authorization, which means remote MCP servers can authenticate using supported MCP auth flows.

About authentication
You may be redirected to Octoparse to sign in and approve access. This is typically handled through an authorization flow rather than by giving VS Code your Octoparse password directly.

Step 6: Use Octoparse in Copilot

Once the Octoparse MCP server is connected, you can use it from GitHub Copilot in VS Code.

Microsoft’s documentation says MCP tools are available in Copilot workflows, including agent mode, where the model can invoke tools from MCP servers to complete tasks.

Try prompts like:

  • Find an Octoparse template for scraping Amazon product listings

  • Create a task using that template

  • Start the task in the cloud

  • Check the status of my latest Octoparse task

  • Export the latest results as CSV

Where Octoparse Works Best in VS Code

Octoparse MCP is most useful in tool-enabled Copilot experiences, especially when you want the model to take multi-step actions.

Microsoft’s documentation says agent mode can use contributed tools, including tools from MCP servers.

Important Usage Notes

  • Some Octoparse actions may require a supported plan or available cloud credits

  • Some tasks may be local-only and not suitable for remote execution through MCP

  • Long, multi-step prompts are more likely to fail than short, focused requests

  • Organization policies in VS Code can restrict MCP usage through admin-managed settings such as chat.mcp.access

Tip
For best results, use shorter commands in sequence:

  1. Find template

  2. Create task

  3. Start task

  4. Check status

  5. Export data

Troubleshooting

If Octoparse does not work in VS Code, check the following:

The MCP server is not loading

Make sure:

  • mcp.json is valid

  • the file is in the correct location

  • VS Code has reloaded the configuration properly

Tools do not appear in Copilot

Make sure:

  • GitHub Copilot is enabled

  • you are using a tool-capable Copilot surface

  • agent mode or the relevant workflow is active when needed

Permissions are blocking usage

Your organization may manage which MCP servers are allowed through settings like chat.mcp.access. If you are in a managed environment, contact your administrator.

Authentication does not complete

Try:

  • reconnecting the server

  • retrying the sign-in flow

  • restarting VS Code after the auth flow finishes

FAQ

Does VS Code support MCP officially?

Yes. Microsoft officially documents adding and managing MCP servers in VS Code through GitHub Copilot.

Where is MCP configuration stored in VS Code?

In mcp.json, either in .vscode/mcp.json for a workspace or in your user profile.

Can VS Code use remote MCP servers?

Yes. VS Code supports MCP server integrations and, as of June 2025, supports the full MCP specification including authorization.

Do I need agent mode to use Octoparse MCP?

Agent mode is one of the main places where MCP tools are used in VS Code, especially for multi-step actions.

Related Articles
How to Integrate Octoparse with ChatGPT via MCP (Step-by-Step)
How to Integrate Octoparse with Claude via MCP (Step-by-Step)
Octoparse MCP Not Working? Common Issues and Fixes
How to Integrate Octoparse with Gemini CLI via MCP (Step-by-Step)
How to Integrate Octoparse with Cursor via MCP (Step-by-Step)
Did this answer your question?