Native MCP

Prioritize MCP-native resources and tools over shell commands.

577|62|Updated May 15, 2026
One-click install
npx skills add https://github.com/agentic-in/elephant-agent --skill native-mcp-agentic-in
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: Native MCP
Source: https://github.com/agentic-in/elephant-agent/tree/main/packages/skills/builtin_packages/mcp/native-mcp
Command: npx skills add https://github.com/agentic-in/elephant-agent --skill native-mcp-agentic-in

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Work through MCP-native resources and tools first when a connector already exposes the needed context or action surface.

Core Features & Use Cases

  • Prefer the connector-native resource or tool over shell scraping.
  • Read structured metadata before falling back to ad hoc commands.
  • Keep the task inside the connector boundary when that preserves correctness and auditability.

Quick Start

Activate the Native MCP flow to leverage built-in MCP datasets and tools before resorting to shell scripting.

Frequently Asked Questions about Native MCP

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

FAQPage Schema
Why should I use MCP-native tools instead of shell scraping for data retrieval?

Use MCP-native tools over shell scraping to ensure safer, traceable data retrieval. This approach reads structured metadata first and keeps tasks within the connector boundary, avoiding brittle shell work.

How do I execute commands through MCP connectors without falling back to shell scripts?

Execute commands by prioritizing connector-native capabilities that expose the needed action surface. This ensures command execution stays within the MCP boundary, using structured tools rather than resorting to ad hoc shell scripting.

What is the best way to handle workflow orchestration when an MCP connector exposes the needed context?

The best way to handle workflow orchestration is to use the native MCP resources and tools. This keeps the workflow inside the connector boundary, preserving correctness and auditability without needing external shell commands.

Does this approach support fallbacks when a native MCP surface is unavailable?

Yes, when a native MCP surface is unavailable, the approach supports documenting fallbacks. It prioritizes reading structured metadata first to find native capabilities, and only falls back to ad hoc commands when a native surface is missing.

When should I not use connector-native tools for my workflow?

You should not use connector-native tools when a connector does not expose the needed context or action surface. In these cases, the task cannot stay within the MCP boundary, requiring documented fallbacks to ad hoc shell commands.