ck:use-mcp

Coordinate MCP server tools via Gemini with stdin piping and fallbacks.

1|Updated Mar 6, 2026
One-click install
npx skills add https://github.com/quanganh208/cookmate --skill ck-use-mcp-quanganh208
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ck:use-mcp
Source: https://github.com/quanganh208/cookmate/tree/main/.opencode/skills/use-mcp
Command: npx skills add https://github.com/quanganh208/cookmate --skill ck-use-mcp-quanganh208

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Coordinate and orchestrate MCP server tools using Gemini to preserve context budget and streamline tool execution across tasks.

Core Features & Use Cases

  • Intelligent discovery and execution: Leverage Gemini and the mcp-manager subagent to locate and run MCP tools without writing new scripts.
  • Safe, constrained execution: All actions follow MCP tooling standards with explicit fallbacks and no script modifications.
  • Use Case: When you need to run a specific MCP operation for a task, the skill orchestrates the appropriate tool, reports results in a structured JSON format, and gracefully falls back if Gemini isn't available.

Quick Start

Provide your task as an argument to Gemini via stdin to execute MCP tools and allow the mcp-manager fallback when needed.

Frequently Asked Questions about ck:use-mcp

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

FAQPage Schema
How do I execute MCP server tools while preserving my LLM context budget?

Coordinate MCP server tools using Gemini to preserve context budget by piping your task via stdin, enabling intelligent tool discovery and execution without writing new scripts.

What happens to MCP tool execution if Gemini is unavailable?

If Gemini is unavailable, MCP tool execution gracefully falls back to the mcp-manager subagent, ensuring your operations continue to run safely within controlled environments.

How do I run MCP operations without writing custom automation scripts?

You can run MCP operations without custom scripts by piping task arguments to Gemini via stdin, which orchestrates the appropriate MCP tools automatically and reports results in structured JSON format.

Do I need to modify existing scripts to use the mcp-manager fallback for tool execution?

No modifications to existing scripts are needed to use the mcp-manager fallback for tool execution, as all actions follow MCP tooling standards with explicit fallbacks and zero script modifications.

What is the best way to orchestrate MCP tools across multiple tasks safely?

The best way to safely orchestrate MCP tools across tasks is using Gemini with mcp-manager fallbacks, applying constrained execution standards that report structured JSON results without script modifications.