Toolport
Open SiteToolport is a free, open-source local MCP gateway that lets AI clients share MCP servers through one setup. It reduces tool-list token bloat, keeps secrets in the OS keychain, and gives developers a cleaner way to use Claude, Cursor, Codex, VS Code, and other agent clients.
Added on July 2, 2026
Product Information
What is Toolport?
Toolport is a local gateway for Model Context Protocol servers. Instead of pasting the same MCP server configuration and API keys into every AI client, developers set each server up once and expose a much smaller tool interface to Claude, Cursor, Codex, VS Code, Windsurf, and other tools. The product focuses on reducing tool-definition tokens, centralizing MCP setup, and keeping secrets on the user's machine. It is useful for developers and agent-heavy teams that rely on multiple AI coding clients but want less configuration drift and context overhead.
How to use Toolport?
- Download and install Toolport for Windows, macOS, or Linux.
- Add MCP servers and store the required credentials locally through Toolport.
- Connect supported AI clients to the single local Toolport gateway.
- Toggle which servers or tools each client can access from the Toolport app.
- Review calls and token savings while agents use a smaller shared tool surface.
Core Features
- Local MCP gateway — Shares configured MCP servers across multiple AI clients from one local app.
- Token reduction — Lets agents load a much smaller tool interface instead of hundreds of raw tool definitions.
- Multi-client support — Works with Claude, Cursor, VS Code, Windsurf, Codex, and other agent clients.
- Secrets in keychain — Keeps API keys and credentials in the operating system keychain rather than copying them into every client.
- Server toggles — Allows users to enable, disable, and route MCP servers without repeatedly editing config files.
- Open-source app — Provides a free and inspectable local workflow with no required cloud account.
Use Cases
- AI coding setup — Configure MCP servers once and reuse them across daily coding assistants.
- Context cleanup — Reduce the tool-definition tokens that slow down or confuse AI agents.
- Team standardization — Give a team a consistent local MCP gateway pattern across supported clients.
- Credential hygiene — Avoid scattering the same server keys across many separate app configs.