Overseer MCP Server

Overseer MCP Server

Provides structured project management through phase-based workflows, enabling planning, execution tracking, compliance checking, and automated documentation for software development projects.

Category
访问服务器

README

Overseer MCP Server

Version: 1.0.0

A standalone Model Context Protocol (MCP) server that implements Overseer multi-agent behavior for structured project management. Overseer provides planning, execution, and enforcement capabilities for managing software projects through well-defined phases.

Quick Start

# Install dependencies
npm install

# Build
npm run build

# Run
npm start

See RUNNING.md for detailed installation and deployment instructions.

Problem Statement

Modern software development involves complex workflows with multiple phases: planning, implementation, testing, deployment, and maintenance. Without structured oversight, projects can:

  • Lose track of progress across multiple workstreams
  • Skip critical steps in development lifecycle
  • Lack visibility into what's been completed vs. what's pending
  • Struggle with consistency across team members and projects
  • Miss documentation and artifact requirements

Overseer exists to solve these problems by:

  1. Enforcing structure through phase-based project management
  2. Tracking progress with clear status indicators and artifacts
  3. Validating completeness before advancing to next phases
  4. Maintaining documentation automatically as projects evolve
  5. Providing tooling that works with any MCP-compatible client

High-Level Capabilities

Planning

  • Project Planning: Create phase definitions from templates or custom specifications
  • Phase Inference: Automatically detect phases from existing project structure
  • Template Management: Use predefined phase templates or create custom ones

Execution

  • Phase Execution: Run specific phases with validation and artifact tracking
  • Phase Advancement: Move phases through lifecycle (pending → active → completed)
  • Status Tracking: Real-time visibility into project and phase status

Enforcement

  • Compliance Checking: Validate that phases meet completion criteria
  • Linting: Ensure code and documentation meet standards
  • Documentation Sync: Keep project docs in sync with actual implementation

Environment & Configuration

  • Environment Mapping: Track and manage environment variables across phases
  • CI/CD Generation: Generate CI/CD pipelines from phase definitions
  • Secrets Management: Create templates for secure credential management

Intended Tech Stack

  • Runtime: Node.js 18+ (ESM modules)
  • Language: TypeScript 5.3+
  • MCP SDK: @modelcontextprotocol/sdk (v0.5.0+)
  • Configuration: YAML (via yaml package)
  • File System: Native Node.js fs/promises
  • Transport: stdio (standard MCP transport)

Architecture Overview

Overseer operates as a pure MCP server with no client-specific dependencies:

┌─────────────────┐
│  MCP Client     │  (Cursor, Claude, Nova, or any MCP client)
│  (any client)   │
└────────┬────────┘
         │ MCP Protocol (stdio/SSE/HTTP)
         │
┌────────▼─────────────────────────────┐
│     Overseer MCP Server              │
│  ┌─────────────────────────────────┐ │
│  │  Tool Layer                     │ │
│  │  (plan_project, run_phase, etc)│ │
│  └────────────┬────────────────────┘ │
│  ┌────────────▼────────────────────┐ │
│  │  Core Logic Layer               │ │
│  │  - PhaseManager                 │ │
│  │  - RepoHandler                  │ │
│  │  - ConfigLoader                 │ │
│  └────────────┬────────────────────┘ │
└───────────────┼──────────────────────┘
                │
┌───────────────▼──────────────────────┐
│  File System                         │
│  - ~/dev/{repo}/PHASES.md           │
│  - ~/dev/{repo}/PHASE-*.md          │
│  - config/sentinel.yml              │
└──────────────────────────────────────┘

Example Use Cases

Use Case 1: Phoenix + Supabase Application

Scenario: Building a full-stack web application with Elixir/Phoenix backend and Supabase frontend.

{
  "repo_name": "phoenix-supabase-app",
  "phases": [
    "planning",
    "database-design",
    "backend-api",
    "frontend-integration",
    "testing",
    "deployment"
  ]
}

Workflow:

  1. overseer.plan_project creates phase structure
  2. overseer.run_phase executes each phase sequentially
  3. overseer.status tracks progress across all phases
  4. overseer.check_compliance validates before deployment
  5. overseer.generate_ci creates CI/CD pipeline

Use Case 2: WordPress Infrastructure Repository

Scenario: Managing infrastructure-as-code for WordPress hosting.

{
  "repo_name": "wordpress-infra",
  "phases": [
    "infrastructure-planning",
    "terraform-setup",
    "kubernetes-config",
    "monitoring-setup",
    "security-hardening",
    "documentation"
  ]
}

Workflow:

  1. overseer.plan_project sets up infrastructure phases
  2. overseer.infer_phases detects existing Terraform/K8s configs
  3. overseer.sync_docs keeps infrastructure docs updated
  4. overseer.env_map tracks environment variables
  5. overseer.secrets_template generates secrets management structure

Use Case 3: Multi-Phase Feature Development

Scenario: Adding a new feature to an existing project.

{
  "repo_name": "existing-project",
  "phases": ["feature-planning", "implementation", "testing", "documentation"]
}

Workflow:

  1. overseer.plan_project adds new phases to existing project
  2. overseer.run_phase executes feature development
  3. overseer.advance_phase moves through lifecycle
  4. overseer.lint_repo ensures code quality
  5. overseer.status provides visibility to team

Installation

See RUNNING.md for detailed installation and setup instructions.

Quick start:

npm install
npm run build
npm start

Configuration

The server reads configuration from config/sentinel.yml. See DESIGN.md for detailed configuration schema.

MCP Client Integration

Cursor IDE

Add to your Cursor MCP configuration (typically in Cursor settings or ~/.cursor/mcp.json):

{
  "mcpServers": {
    "overseer": {
      "command": "node",
      "args": ["/absolute/path/to/overseer-mcp/dist/server.js"],
      "env": {
        "OVERSEER_BASE_PATH": "~/dev"
      }
    }
  }
}

Using Docker:

{
  "mcpServers": {
    "overseer": {
      "command": "docker",
      "args": [
        "run", "-i", "--rm",
        "-v", "~/dev:/root/dev:ro",
        "-v", "/absolute/path/to/overseer-mcp/config:/app/config:ro",
        "freqkflag/overseer-mcp:latest"
      ]
    }
  }
}

Claude Desktop

Add to ~/Library/Application Support/Claude/claude_desktop_config.json (macOS):

{
  "mcpServers": {
    "overseer": {
      "command": "node",
      "args": ["/absolute/path/to/overseer-mcp/dist/server.js"]
    }
  }
}

Usage Example

Once configured, you can use Overseer tools in your MCP client:

Plan a project:

{
  "repo_root": "~/dev/sample-project",
  "project_name": "sample-project",
  "project_summary": "A sample web application",
  "overwrite_existing": false
}

Run a phase:

{
  "repo_root": "~/dev/sample-project",
  "phase_id": "01",
  "aggression_level": "normal"
}

Advance phase:

{
  "repo_root": "~/dev/sample-project",
  "expected_current_phase": "01"
}

See DEMO.md for a complete walkthrough.

Available Tools

See TOOLS.md for complete tool documentation.

Project Structure

overseer-mcp/
├── config/
│   └── sentinel.yml          # Configuration file
├── src/
│   ├── core/
│   │   ├── config.ts        # Configuration loader
│   │   ├── phase-manager.ts # Phase management logic
│   │   ├── repo.ts          # Repository file operations
│   │   ├── repo-analyzer.ts # Repository structure analysis
│   │   └── fsUtils.ts       # File system utilities
│   ├── tools/
│   │   ├── plan-project.ts  # Project planning
│   │   ├── infer-phases.ts  # Phase inference
│   │   ├── update-phases.ts # Phase updates
│   │   ├── run-phase.ts     # Phase execution
│   │   ├── advance-phase.ts # Phase advancement
│   │   ├── status.ts         # Project status
│   │   ├── lint-repo.ts     # Repository linting
│   │   ├── sync-docs.ts     # Documentation sync
│   │   ├── check-compliance.ts # Compliance checking
│   │   ├── env-map.ts       # Environment mapping
│   │   ├── generate-ci.ts   # CI/CD generation
│   │   ├── secrets-template.ts # Secrets templates
│   │   └── index.ts         # Tool registration
│   └── server.ts            # MCP server entry point
├── config/
│   └── sentinel.yml         # Configuration
├── Dockerfile               # Docker image definition
├── docker-compose.yml      # Docker Compose configuration
├── package.json
├── tsconfig.json
├── README.md                # This file
├── RUNNING.md               # Installation and usage guide
├── DEMO.md                  # Demo scenario walkthrough
├── DESIGN.md                # Architecture and design
├── TOOLS.md                 # Tool documentation
└── PHASES.md                # Build phases for this project

Development

See RUNNING.md for detailed instructions.

Quick commands:

# Build
npm run build

# Development mode with watch
npm run dev

# Run production server
npm start

# Docker
docker-compose up -d

Client-Agnostic Design

Overseer is designed to work with any MCP-compatible client:

  • Cursor: IDE integration via MCP
  • Claude Desktop: Chat-based interaction
  • Nova: Code editor integration
  • Custom clients: Any tool that speaks MCP protocol

All tool interfaces use pure JSON-compatible structures. No client-specific features are required.

License

MIT

推荐服务器

Baidu Map

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。

官方
精选
JavaScript
Playwright MCP Server

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。

官方
精选
TypeScript
Magic Component Platform (MCP)

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。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

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。

官方
精选
本地
TypeScript
VeyraX

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。

官方
精选
本地
graphlit-mcp-server

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。

官方
精选
TypeScript
Kagi MCP Server

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。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

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。

官方
精选