OmniPlan MCP Server

OmniPlan MCP Server

Enables Claude to read and analyze project schedule files in OmniPlan (.oplx) and Microsoft Project (.mpp) formats, answering questions about tasks, milestones, resources, and progress.

Category
访问服务器

README

<p align="center"> <img src="https://img.shields.io/badge/macOS-required-blue" alt="macOS"> <img src="https://img.shields.io/badge/python-≥3.10-blue" alt="Python"> <img src="https://img.shields.io/badge/license-MIT-green" alt="License"> <img src="https://img.shields.io/github/v/release/cygnusyang/omniplan-mcp" alt="Release"> </p>

OmniPlan MCP Server

A Model Context Protocol (MCP) server that lets Claude read and analyze project schedule files — OmniPlan (.oplx) and Microsoft Project (.mpp) formats.

Ask Claude questions like:

  • "What's the current project schedule?"
  • "List all milestones and their dates"
  • "Show me tasks related to the robotic arm"
  • "What's the overall progress percentage?"

Features

Feature Description
📂 Read .mpp Parse Microsoft Project files via OmniPlan bridge
📂 Read .oplx Direct XML parsing (no OmniPlan needed)
🏛️ Full hierarchy Groups, tasks, milestones with dates and progress
🔍 Search Find tasks by keyword across the entire schedule
👤 Resources List all human resources and assignments
📊 Summary Phase overview, progress statistics, timeline
🔒 Safe concurrency Direct AppleScript reading avoids temp-file conflicts when multiple sessions run

Prerequisites

Requirement Notes
macOS Required (for AppleScript/OmniPlan bridge)
Python 3.10+ For running the MCP server
OmniPlan Only needed for .mpp files; .oplx works without it

Install OmniPlan (optional — only for .mpp files)

brew install --cask omniplan

First run: macOS may prompt for Accessibility/Automation permissions when OmniPlan is called via AppleScript. Grant them in System Settings → Privacy & Security → Automation.

Quick Start

1. Install

# Option A: One-line installer (recommended)
curl -fsSL https://raw.githubusercontent.com/cygnusyang/omniplan-mcp/main/install.sh | bash

# Option B: Manual clone
git clone https://github.com/cygnusyang/omniplan-mcp.git
cd omniplan-mcp
pip install -e .

2. Configure Claude Code

Add to your ~/.claude/settings.json:

<details> <summary><b>uv run (recommended)</b></summary>

{
  "mcpServers": {
    "omniplan": {
      "command": "uv",
      "args": [
        "run",
        "--directory", "/Users/yourusername/.local/share/omniplan-mcp",
        "omniplan-mcp"
      ],
      "env": {}
    }
  }
}

</details>

<details> <summary><b>pip install (after PyPI publish)</b></summary>

{
  "mcpServers": {
    "omniplan": {
      "command": "uvx",
      "args": ["omniplan-mcp"],
      "env": {}
    }
  }
}

</details>

<details> <summary><b>Direct Python</b></summary>

{
  "mcpServers": {
    "omniplan": {
      "command": "/path/to/python3",
      "args": ["-m", "omniplan_mcp"],
      "env": {
        "PYTHONPATH": "/path/to/omniplan-mcp/src"
      }
    }
  }
}

</details>

3. Restart Claude Code

The MCP server will start automatically. You can now ask Claude about your project files!

Usage Examples

Read a project schedule

你:帮我读取 PLB1011 项目计划,看看有哪些阶段
Claude:调用 read_schedule → 显示完整任务树

List milestones

你:列出所有里程碑节点
Claude:调用 list_milestones → 显示所有 ◇ 里程碑

Search for tasks

你:搜索所有关于"机械臂"的任务
Claude:调用 search_tasks → 显示匹配的任务列表

Project summary

你:这个项目的整体进度怎么样?
Claude:调用 schedule_summary → 显示阶段概览和进度统计

Tools Reference

Tool Description Parameters
read_schedule Full task hierarchy with dates and progress filepath (required), format: tree/flat/json
list_milestones All milestone tasks filepath
list_resources All human resources filepath, detail: simple/full
search_tasks Search tasks by keyword filepath, keyword
schedule_summary Phase overview and progress stats filepath
get_task_detail Detailed info about a specific task filepath, task_id or task_name
get_resource_detail Detailed info about a specific resource filepath, resource_name
list_violations All scheduling conflicts/violations filepath
list_assignments All resource-to-task assignments filepath
list_dependencies All task dependency relationships filepath
get_schedule_settings Scheduling granularity & working hours (reads active OmniPlan document)
evaluate_omniplan_script Run Omni Automation JS in OmniPlan script (JavaScript code)
export_schedule Export schedule to various formats filepath, format (optional), output_path (optional)
lookup_task Find task by name, get its numeric ID search_name
set_task_completed Set task to 100% complete task_id, include_subtree
set_task_completed_by_name Set task to 100% complete by name task_name, include_subtree
add_dependency Add finish-to-start dependency dependent_task_id, prerequisite_task_id
remove_dependency Remove a dependency dependent_task_id, prerequisite_task_id
set_task_duration Change task duration task_id, duration_seconds
clear_constraint_date Remove locked start date task_id
rename_task Rename a task task_id, new_name
delete_task Delete a task and its children task_id
add_task Add a new task under a parent parent_task_id, task_name, duration_seconds (optional)
save_document Save the OmniPlan document (none)

New in v0.4.0

  • 12 new write-operation tools — Now you can modify schedules directly from Claude:
    • lookup_task — Find any task by name to get its numeric ID
    • set_task_completed / set_task_completed_by_name — Mark tasks as 100% complete
    • add_dependency / remove_dependency — Manage task dependencies (prerequisites)
    • set_task_duration — Adjust task durations
    • clear_constraint_date — Remove locked/constraint dates
    • rename_task / delete_task / add_task — Structure editing
    • save_document — Persist changes to disk
  • Bug fixes: .oplx parsing now prefers Actual.xml (fixes stale backup reads), outline_depth computed from hierarchy, percent-complete derived from effort-done/effort ratio, task_status computed for .oplx tasks, evaluate_javascript quotes properly escaped
  • Parser robustness: build_task_tree handles both string and integer parent_ids

How It Works

.mpp file ──→ OmniPlan (AppleScript direct read) ──→ pipe-delimited records ──→ Claude
                          ↑
.oplx file ───────────────┴─── direct XML parsing ──────┘

For .oplx files

Direct XML parsing — fast, no external dependencies.

For .mpp files

  1. MCP server opens the .mpp file in OmniPlan via the macOS open command
  2. Reads all project data (tasks, resources, dates, progress) directly from OmniPlan's in-memory object model via AppleScript
  3. Parses the pipe-delimited output into structured records
  4. Closes the document

No temporary files are created — data is read directly from OmniPlan's in-memory model.

Project Structure

omniplan-mcp/
├── install.sh                  # One-click installer
├── pyproject.toml              # Package metadata (PyPI-ready)
├── README.md                   # This file
├── LICENSE                     # MIT license
├── .gitignore
├── src/
│   └── omniplan_mcp/
│       ├── __init__.py         # Package version
│       ├── __main__.py         # CLI entry point
│       ├── server.py           # MCP server (tools & handlers)
│       └── parser.py           # .mpp (AppleScript) / .oplx (XML) parsing
└── tests/
    └── test_parser.py          # Unit tests

Development

# Clone
git clone https://github.com/cygnusyang/omniplan-mcp.git
cd omniplan-mcp

# Install in editable mode
pip install -e .

# Run tests
python -m pytest tests/

# Run the server directly (stdio)
python -m omniplan_mcp

Publishing to PyPI

Published automatically via GitHub Actions (Trusted Publisher) when a tag is pushed:

git tag v0.1.0
git push origin v0.1.0

Manual build (for testing):

pip install build
python -m build

Requirements

  • Python 3.10+
  • macOS (for OmniPlan AppleScript bridge)
  • OmniPlan (only for .mpp files; optional for .oplx)

Limitations

  • .mpp parsing requires OmniPlan to be installed
  • Only supports macOS (AppleScript dependency)
  • Does not modify .mpp files — read-only

License

MIT License — see LICENSE for details.

Related

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