native-mcp

Integrate external MCP servers and their tools via stdio or HTTP transports.

2|7|Updated Jun 19, 2026
One-click install
npx skills add https://github.com/humanerd-drew/opencode-drewgent --skill native-mcp-humanerd-drew
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: native-mcp
Source: https://github.com/humanerd-drew/opencode-drewgent/tree/main/skills/mcp/native-mcp
Command: npx skills add https://github.com/humanerd-drew/opencode-drewgent --skill native-mcp-humanerd-drew

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill Unit allows the AI agent to connect to external MCP servers, automatically discover and integrate their tools, and use them as if they were built-in features.

Core Features & Use Cases

  • External Tool Integration: Seamlessly add capabilities like filesystem access, GitHub, databases, and APIs.
  • stdio and HTTP Transports: Supports both command-based and URL-based server connections.
  • Automatic Reconnection: Ensures continuous availability of MCP tools, even if the connection drops.
  • Tool Naming Convention: MCP tools are registered with a clear naming pattern for easy identification.
  • Security Considerations: Environment variable filtering and credential wrapping to protect sensitive information.

Quick Start

Configure MCP servers in your agent's configuration file, and then use them in conversations like this: "Get the current time from the 'time' server."

Frequently Asked Questions about native-mcp

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

FAQPage Schema
How do I integrate external MCP servers into an AI agent?

Configure external MCP servers in your agent's configuration file to automatically discover and integrate their tools into the AI agent environment. This supports both stdio and HTTP connection types.

What is the difference between stdio and HTTP transports for MCP tool integration?

Stdio transport connects to MCP servers via command-based execution, while HTTP transport uses URL-based server connections. Both connection types allow the AI agent to access external resources and capabilities seamlessly.

Can I connect to filesystem, GitHub, and database APIs using MCP servers?

Yes, connecting external MCP servers allows you to seamlessly add capabilities like filesystem access, GitHub integration, databases, and APIs to your AI agent. These external tools are integrated automatically upon discovery.

How are credentials and environment variables protected when connecting to MCP servers?

Credential security during MCP server integration is handled through environment variable filtering and credential wrapping. This protects sensitive information while the agent accesses external resources.

Does the MCP connection automatically recover if the external server drops?

Yes, automatic reconnection ensures the continuous availability of MCP tools even if the connection drops. This maintains uninterrupted access to external resources and capabilities for the AI agent.

How are external MCP tools named when registered in the AI agent?

MCP tools are registered with a clear naming convention pattern for easy identification within the agent environment. This allows seamless use of external tools in conversations, such as querying a time server.