AI Client Integration (IDE Integration)
ServBay lets you quickly integrate its built-in MCP Server with your favorite AI clients in just one click, automatically updating the relevant MCP configurations—no manual editing required. This article describes how to integrate, which clients are supported, and how each integration works.
Prerequisites
- ServBay is installed and running.
- The target AI client is installed (ServBay checks its own configuration to determine if integration is possible).
One-Click Integration
Navigate to ServBay Settings → AI Client Integrations (IDE Integrations). Find your target client in the list and click Integrate. ServBay will:
- Locate the bundled
servbay-mcpexecutable (no separate installation needed). - Update the MCP configuration according to the target client's specification (only adds/updates the
servbaynode—your existing MCP configurations are preserved). - Display the integration status.
To disconnect, simply click Remove.
Integration Status
Each client displays its integration status relative to "this local instance of ServBay":
| Status | Meaning |
|---|---|
| Integrated | Client points to this ServBay's local servbay-mcp |
| Integrated (Other Path) | Client has servbay configured but points elsewhere (such as another ServBay instance) |
| Not Integrated | Client detected, but not yet integrated |
| Not Detected | This client was not found on your machine |
Supported Clients and Integration Methods
| Client | Integration Method | Configuration Location |
|---|---|---|
| Claude Code | Official CLI: claude mcp add --scope user --transport stdio servbay -- <bin> | Claude Code user-level MCP config |
| Codex | Official CLI: codex mcp add (falls back to writing ~/.codex/config.toml if CLI isn't available) | Codex config |
| Cursor | Write ~/.cursor/mcp.json (top-level mcpServers) | ~/.cursor/mcp.json |
| VS Code / GitHub Copilot | Write user-level mcp.json (top-level servers, type: stdio) | VS Code user-level mcp.json |
| Claude Desktop | Write claude_desktop_config.json (top-level mcpServers) | claude_desktop_config.json |
MCP configuration schema differences
The top-level key for MCP configuration varies by client: VS Code uses servers, while Cursor and Claude Desktop use mcpServers. ServBay automatically writes the correct key for each client, so you don't need to worry about these differences.
Verifying Integration
After integration, initiate an environment-related request from the AI client (for example, "List the currently running services in ServBay"). If the AI returns an actual list of services, the integration was successful.
Frequently Asked Questions (FAQ)
- Q: Will integration overwrite my existing MCP configurations?
- A: No. ServBay only adds or updates the
servbaynode, preserving your other configured MCP Servers.
- A: No. ServBay only adds or updates the
- Q: Why does the client show 'Not Detected'?
- A: This means the client was not found on your computer. Please install and launch the client at least once before integrating.
- Q: The AI fails to connect after integration.
- A: Make sure ServBay is running; for Claude Code and Codex (CLI integration), ensure their official CLIs are available. Try removing and re-integrating if necessary.
- Q: What is recommended after integration?
- A: We recommend also installing AI Rules to guide the AI to use ServBay MCP first and correctly.
Summary
With "AI Client Integration," you can seamlessly connect ServBay MCP to the most popular AI tools with a single click—no manual configuration and no disruption to your current setup. Once integrated, pairing with AI Rules lets your AI effortlessly manage your local development environment.
