Перейти к содержимому
TakoAPI
Назад к навыкам

mcp-workflow

от slemo54

Productivity & Tasks
442Число загрузок по данным clawskills.sh
0Просмотры
0 лайковВойдите, чтобы поставить лайк навыку
—Ещё нет оценок
Ваша оценка

Кратко

Workflow automation using MCP (Model Context Protocol) patterns inspired by Jason Zhou.

Описание

Что делает этот навык

Runs multi-step AI workflows using MCP (Model Context Protocol), coordinating prompt chains, resource loading, and cross-server execution. Includes built-in templates for common tasks like code review, meal planning, and report generation. Workflows are defined as JSON and run through a local MCP server.

MCP's standardized resource URIs and tool interfaces let workflows coordinate across multiple AI servers without custom glue code for each integration.

Когда его использовать

  • Generating a weekly team status report from project metrics
  • Running automated code review on a pull request
  • Building a 7-day meal plan with a shopping list
  • Producing API documentation from source files
  • Chaining multi-step prompts for structured data processing

Пример сценария

ВВОД

User asks: Create a 7-day vegetarian meal plan with a $50 budget

АГЕНТ
  1. 1

    Loads the meal-planner workflow template

  2. 2

    Validates input parameters against the workflow schema

  3. 3

    Executes the prompt chain: plan → generate → execute → validate

  4. 4

    Embeds file and memory resources as needed at each step

ВЫВОД

A 7-day vegetarian meal plan with a categorized shopping list under the specified budget

Установка

clawhub install slemo54/mcp-workflow

Выполните команду CLI или вставьте промпт в Claude Code, Cursor, Codex — любой кодинг-агент, которым пользуетесь.

Ссылки