---
title: "Use Langfuse in VS Code via MCP"
sidebarTitle: VS Code
logo: /images/integrations/vscode_icon.svg
description: "Connect the Langfuse MCP server to VS Code to query and manage your prompts, traces, and datasets from GitHub Copilot agent mode."
category: Integrations
---

# Use Langfuse in VS Code via MCP

This guide shows you how to connect the [Langfuse MCP server](/docs/api-and-data-platform/features/mcp-server) to [VS Code](https://code.visualstudio.com), so you can query and manage your Langfuse data directly from GitHub Copilot's agent mode.

> **What is VS Code?** [**VS Code**](https://code.visualstudio.com) is a code editor from Microsoft with built-in AI assistance through GitHub Copilot. Its [agent mode](https://code.visualstudio.com/docs/copilot/chat/mcp-servers) supports the Model Context Protocol (MCP), letting Copilot call external tools while it helps you write and debug code.

> **What is Langfuse?** [**Langfuse**](https://langfuse.com) is the open-source AI engineering platform. It helps teams collaboratively debug, analyze, and iterate on their LLM applications with features like tracing, prompt management, and evaluation.

## What This Integration Does

Langfuse ships a native [Model Context Protocol (MCP) server](/docs/api-and-data-platform/features/mcp-server) for the data platform. Once connected to VS Code, Copilot agent mode can use it to:

- **Manage prompts**: List, read, create, and update prompts in [Langfuse Prompt Management](/docs/prompt-management/overview) without leaving your editor.
- **Inspect traces**: Pull traces and observations to debug your LLM application against real production data.
- **Work with datasets**: Browse datasets and items used for evaluation.
- **Access the platform programmatically**: Reach the same resources exposed by the Langfuse [Public API](/docs/api-and-data-platform/features/public-api).

This page covers the **authenticated** Langfuse MCP server for the data platform. There is also a public [Docs MCP server](/docs/docs-mcp) that exposes the Langfuse documentation to AI agents — useful in VS Code for [automatically integrating Langfuse tracing](/docs/get-started) into your codebase.

## Prerequisites

- [VS Code](https://code.visualstudio.com) 1.99 or later with [GitHub Copilot](https://code.visualstudio.com/docs/copilot/setup) enabled
- A [Langfuse](https://langfuse.com) account ([cloud](https://cloud.langfuse.com) or [self-hosted](/self-hosting))

## Quick Start

<Steps>

### Get an Authentication Header

1. Navigate to your project settings in Langfuse and create or copy a **project-scoped API key**:
   - Public Key: `pk-lf-...`
   - Secret Key: `sk-lf-...`
2. Encode the credentials to base64 format:

   ```bash filename="terminal"
   echo -n "pk-lf-your-public-key:sk-lf-your-secret-key" | base64
   ```

   You'll use the resulting token as `{your-base64-token}` below.

### Add the MCP Server to VS Code

Create a `.mcp.json` file in your workspace's `.vscode` folder (or add the server via the command palette: `MCP: Add Server...` → `HTTP`). Replace `{your-base64-token}` with your encoded credentials and pick the endpoint for your data region:

<Tabs items={["Cloud EU", "Cloud US", "Cloud Japan", "HIPAA US", "Self-Hosted"]}>

<Tab>

```json filename=".vscode/mcp.json" /{your-base64-token}/
{
  "servers": {
    "langfuse": {
      "type": "http",
      "url": "https://cloud.langfuse.com/api/public/mcp",
      "headers": {
        "Authorization": "Basic {your-base64-token}"
      }
    }
  }
}
```

</Tab>

<Tab>

```json filename=".vscode/mcp.json" /{your-base64-token}/
{
  "servers": {
    "langfuse": {
      "type": "http",
      "url": "https://us.cloud.langfuse.com/api/public/mcp",
      "headers": {
        "Authorization": "Basic {your-base64-token}"
      }
    }
  }
}
```

</Tab>

<Tab>

```json filename=".vscode/mcp.json" /{your-base64-token}/
{
  "servers": {
    "langfuse": {
      "type": "http",
      "url": "https://jp.cloud.langfuse.com/api/public/mcp",
      "headers": {
        "Authorization": "Basic {your-base64-token}"
      }
    }
  }
}
```

</Tab>

<Tab>

```json filename=".vscode/mcp.json" /{your-base64-token}/
{
  "servers": {
    "langfuse": {
      "type": "http",
      "url": "https://hipaa.cloud.langfuse.com/api/public/mcp",
      "headers": {
        "Authorization": "Basic {your-base64-token}"
      }
    }
  }
}
```

</Tab>

<Tab>

```json filename=".vscode/mcp.json" /{your-base64-token}/
{
  "servers": {
    "langfuse": {
      "type": "http",
      "url": "https://your-domain.com/api/public/mcp",
      "headers": {
        "Authorization": "Basic {your-base64-token}"
      }
    }
  }
}
```

</Tab>

</Tabs>

To make the server available across all your workspaces, add the same `langfuse` entry to your user-level MCP configuration instead via `MCP: Open User Configuration` from the command palette.

### Start the Server

Open the `.vscode/mcp.json` file and click **Start** above the `langfuse` server entry, or run `MCP: List Servers` from the command palette and start it from there. VS Code will connect to the Langfuse MCP server.

### Verify the Connection

1. Open the Copilot Chat view and switch to **Agent** mode.
2. Confirm the Langfuse tools are available via the **Tools** picker.
3. Ask Copilot to `list all prompts in my Langfuse project`. It should call the `listPrompts` tool and return your prompts.

</Steps>

  **Both read and write tools are available by default.** If you only want to
  use read-only tools, restrict the available tools from VS Code's **Tools**
  picker in agent mode. For a full list of tools, see the [MCP Reference](https://mcp.reference.langfuse.com).

## Example Prompts

Once connected, try asking Copilot agent mode:

- "List all prompts in my Langfuse project and show me the latest version of the `chat-assistant` prompt."
- "Create a new text prompt called `summarizer` with the `production` label."
- "Update the labels on the `chat-assistant` prompt to promote version 3 to production."

## Resources

- [Langfuse MCP Server](/docs/api-and-data-platform/features/mcp-server) — full configuration reference for the authenticated MCP server
- [MCP Reference](https://mcp.reference.langfuse.com) — canonical source for MCP servers, setup snippets, tools, and schemas
- [Docs MCP Server](/docs/docs-mcp) — public MCP server for the Langfuse documentation
- [VS Code MCP Servers Documentation](https://code.visualstudio.com/docs/copilot/chat/mcp-servers) — Microsoft's guide to using MCP servers in VS Code
- [Langfuse for Agents](/agents) — overview of the skill, CLI, and MCP server for coding agents

<!-- agent-instructions -->

---

## Agent Instructions

This page is part of the [Langfuse](https://langfuse.com) documentation, published as plain Markdown for AI agents. Every page is available as Markdown by appending `.md` to its URL, or by sending an `Accept: text/markdown` header. This page: `https://langfuse.com/integrations/developer-tools/vscode.md`.

### Querying these docs

If the answer is not on this page, query the documentation instead of guessing:

- **Semantic search** across all Langfuse docs, returning an answer with the relevant pages and excerpts. Ask a specific, self-contained question:

  ```bash
  curl -sG "https://langfuse.com/api/search-docs" --data-urlencode "query=How do I trace a LangGraph agent?"
  ```

- **Index of every page**: <https://langfuse.com/llms.txt>, with per-section indexes [llms-docs.txt](https://langfuse.com/llms-docs.txt), [llms-integrations.txt](https://langfuse.com/llms-integrations.txt), and [llms-self-hosting.txt](https://langfuse.com/llms-self-hosting.txt).

### Before writing Langfuse code

- **Install the [Langfuse Agent Skill](https://langfuse.com/docs/api-and-data-platform/features/agent-skill).** It encodes Langfuse's own best practices for instrumentation, prompt management, and evaluation, and materially improves results.
- **Read [What does a good trace look like?](https://langfuse.com/docs/observability/best-practices.md)** before instrumenting an application.
- **Verify endpoints, parameters, and response fields** against the [API reference](https://api.reference.langfuse.com) instead of inferring them from code examples.
- **Use the [Langfuse CLI](https://langfuse.com/docs/api-and-data-platform/features/cli)** (`npx langfuse-cli api <resource> <action>`) to read or write traces, prompts, datasets, and scores from the terminal.

Found an error in these docs? Please open an issue at <https://github.com/langfuse/langfuse-docs/issues>.