native-mcp

Connect to external MCP servers and expose their tools automatically.

Updated Apr 12, 2026
One-click install
npx skills add https://github.com/DaddyElonMusk69/motis-agent --skill native-mcp-daddyelonmusk69
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: native-mcp
Source: https://github.com/DaddyElonMusk69/motis-agent/tree/main/skills/mcp/native-mcp
Command: npx skills add https://github.com/DaddyElonMusk69/motis-agent --skill native-mcp-daddyelonmusk69

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It eliminates the need for manual bridge scripts by automatically discovering and exposing tools from external MCP servers, letting agents call them as native capabilities.

Core Features & Use Cases

  • Transport Support: Works with both stdio (command‑line) and HTTP transports for diverse server setups.
  • Automatic Registration: Discovered tools are injected into all Motis toolsets with a mcp_{server}_{tool} naming convention.
  • Security Filtering: Passes only safe environment variables to subprocesses and redacts credentials in error messages.
  • Reliability: Persistent connections with exponential backoff and graceful shutdown handling.
  • Use Cases: Connect to a time‑server for current timestamps, a filesystem server for file operations, or a remote API server for data retrieval without writing custom adapters.

Quick Start

Tell Motis to load the native MCP client and then invoke a tool such as mcp_time_get_current_time.

Frequently Asked Questions about native-mcp

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I expose external MCP server tools to my agent without writing manual bridge scripts?

You can expose external MCP server tools without manual bridge scripts by using an auto-discovery client that automatically registers them as native capabilities. This eliminates custom adapters by directly injecting discovered tools into the agent's toolset.

Does MCP client support both stdio command-line and HTTP transports for remote tool access?

The MCP client supports both stdio command-line and HTTP transports for remote tool access. This allows diverse server setups across development and automation workflows to connect without requiring transport-specific adapters.

How are discovered MCP tools named when they are registered with the agent?

Discovered MCP tools are registered with the agent using a specific naming convention. They are automatically injected into all toolsets with the format mcp_{server}_{tool}, ensuring clear identification and immediate usability.

How does the MCP client handle security and credentials in environment variables?

The MCP client handles security by filtering environment variables, passing only safe ones to subprocesses. It also redacts credentials in error messages to prevent sensitive data exposure during remote tool execution.

What happens when an MCP server connection drops during an automation workflow?

When an MCP server connection drops, the client maintains reliability through persistent connections with exponential backoff. It includes graceful shutdown handling to ensure automation workflows recover without manual intervention.

Can I connect to a remote API server for data retrieval using auto discovery?

You can connect to a remote API server for data retrieval using auto discovery. The client automatically discovers and exposes the server's tools, allowing agents to retrieve data without writing custom adapters.