MCP Planning Server
Provides structured task planning and tracking capabilities with 40 tools for managing requirements, solutions, architectural decisions, and hierarchical implementation phases with progress tracking.
README
MCP Planning Server
A Model Context Protocol (MCP) server for Claude Code that provides structured task planning and tracking capabilities.
Features
- 40 planning tools organized into 8 categories
- File-based JSON storage with atomic writes
- Hierarchical phases with progress tracking
- Requirement traceability from definition to implementation
- Decision records (ADR-style) with supersede support
- Solution comparison with tradeoff analysis
Installation
npm install
npm run build
Configuration
Add to your Claude Code MCP settings:
{
"mcpServers": {
"planning": {
"command": "node",
"args": ["path/to/planner/dist/index.js"],
"env": {
"MCP_PLANNING_STORAGE_PATH": "./.mcp-plans"
}
}
}
}
Disabling Permission Prompts
By default, Claude Code asks for confirmation before each MCP tool use. To allow all planning tools without prompts, add to your settings:
{
"permissions": {
"allow": [
"mcp__planning"
]
}
}
Settings locations (in order of priority):
- Project-local:
.claude/settings.local.json— for this project only, not committed to git - Project-shared:
.claude/settings.json— shared with team via git - Global:
~/.claude/settings.json— applies to all your projects
Note: MCP permissions do NOT support wildcards. Use mcp__planning (not mcp__planning__*) to allow all tools from this server.
Usage
Creating a Plan
Simply describe your task to Claude Code. The assistant will automatically create a plan:
"I need to build a user authentication system with OAuth support"
Claude Code will:
- Create a new plan
- Add requirements based on your description
- Propose solutions with tradeoffs
- Break down implementation into phases
Tracking Progress
As you work, Claude Code updates the plan:
"I've finished implementing the login form"
The assistant will mark the corresponding phase as complete and suggest next actions.
Viewing Plan Status
Ask about your current progress:
"What's the status of my plan?" "What should I work on next?"
Tool Categories
| Category | Tools | Description |
|---|---|---|
| Plan Management | 7 | Create, list, get, update, archive plans |
| Requirements | 5 | Manage requirements with priorities and acceptance criteria |
| Solutions | 6 | Propose, compare, and select solutions |
| Decisions | 4 | Record architectural decisions (ADR) |
| Phases | 6 | Hierarchical implementation phases |
| Linking | 3 | Entity relationships with cycle detection |
| Query | 4 | Search, trace, validate, export |
| System | 1 | Health check |
Storage
Plans are stored as JSON files in the configured storage path:
.mcp-plans/
├── plans/
│ └── {plan-id}/
│ ├── manifest.json
│ ├── requirements.json
│ ├── solutions.json
│ ├── decisions.json
│ ├── phases.json
│ └── links.json
└── active-plans.json
Development
# Run tests
npm test
# Build
npm run build
# Type check
npx tsc --noEmit
License
MIT
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。