Accordo
A YAML-driven workflow guidance MCP server that enables AI coding agents to follow structured development workflows with real-time state tracking and progression control.
README
<div align="center"> <img src="logo.png" alt="Accordo Logo" width="1000"/> </div>
Accordo MCP Server
A powerful MCP (Model Context Protocol) server that provides dynamic YAML-driven workflow guidance for AI coding agents. Features structured development workflows with progression control and decision points.
What This Does
This server guides AI agents through structured, schema-driven development workflows:
- 📋 Dynamic YAML Workflows: Custom workflows defined in YAML with schema-driven execution
- 🔍 Discovery-First: Automatically discovers and selects appropriate workflows based on task
- ⚡ Real-time State: Live workflow state tracking with comprehensive session management
- 🎯 Mandatory Guidance: Authoritative phase-by-phase instructions agents must follow
- 🤖 Agent-Generated Workflows: Unique capability for agents to create custom YAML workflows on-the-fly
Two Workflow Approaches
Built-in Workflow Discovery Process
- Automatic Discovery - System scans
.accordo/workflows/for available YAML workflows - Intelligent Matching - Analyzes task description against workflow capabilities and goals
- Workflow Selection - Agent chooses the most appropriate workflow from discovered options
- Immediate Execution - Selected workflow starts with full state management and progression control
This provides instant access to proven workflows for common development patterns - coding, documentation, debugging, and more!
Agent Workflow Generation Process
- Analyze Task Requirements - Agent understands specific needs
- Get Creation Guidance - Comprehensive templates and best practices provided
- Generate Custom YAML - Agent creates workflow tailored to exact requirements
- Start Execution - Custom workflow runs with full state management
This enables agents to create workflows for any domain - web development, data science, DevOps, research, documentation, testing, and more!
Quick Start
Installing accordo CLI
The accordo CLI provides easy configuration and management of MCP servers. Install it directly with:
# Download and run the installation script
curl -fsSL https://raw.githubusercontent.com/AndurilCode/accordo/refs/heads/main/install.sh | bash
The installer will:
- ✅ Detect your environment (virtual env, available tools)
- 🌍 Global installation (recommended): Uses pipx for system-wide access
- 📦 Virtual environment: Uses uv or pip if you're in a venv
- 🔧 Auto-setup: Handles dependencies and PATH configuration
CLI Usage
# Configure MCP servers interactively
accordo configure
# Quick setup for specific platforms
accordo configure -p cursor -y
accordo configure -p claude-code -y
# Deploy workflow guidelines to AI assistants
accordo bootstrap-rules # All assistants
accordo bootstrap-rules cursor # Cursor only
accordo bootstrap-rules --force all # Overwrite existing
# List supported platforms and manage servers
accordo list-platforms
accordo list-servers -p cursor
accordo validate -p cursor
# Get help
accordo --help
Manual MCP Configuration
Alternatively, install uv for Python package management and configure your MCP client manually:
For Cursor (.cursor/mcp.json):
{
"mcpServers": {
"accordo": {
"command": "uvx",
"args": ["accordo-workflow-mcp"]
}
}
}
Or use the following link to install the MCP server (Cursor Installed):
Getting Help
📚 Documentation
- Complete Documentation - Comprehensive guides and references
- Installation Prerequisites - Detailed setup requirements
🎯 Examples
- Workflow Examples - Ready-to-use workflow YAML files
- Configuration Examples - MCP client setup examples
License
This project is licensed under the MIT License - see the LICENSE file for details.
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。