home / skills / openclaw / skills / office-mcp

This skill helps automate Office document tasks by creating and editing Word, Excel, and PowerPoint files via MCP tools.

npx playbooks add skill openclaw/skills --skill office-mcp

Review the files below or copy the command above to add this skill to your agents.

Files (2)
SKILL.md
2.6 KB
---
name: office-mcp
description: MCP server for Word, Excel, PowerPoint operations via AI
author: claude-office-skills
version: "1.0"
tags: ['mcp', 'office', 'word', 'excel', 'powerpoint']
models: [claude-sonnet-4, claude-opus-4]
tools: [computer, code_execution, file_operations]
library:
  name: Office MCP
  url: https://github.com/anthropics/skills
  stars: N/A
---

# Office Mcp Skill

## Overview

This skill wraps Office document operations as MCP tools, allowing Claude to create, edit, and manipulate Word, Excel, and PowerPoint files with standardized interfaces.

## How to Use

1. Describe what you want to accomplish
2. Provide any required input data or files
3. I'll execute the appropriate operations

**Example prompts:**
- "Create Word documents with AI-generated content"
- "Build Excel spreadsheets with formulas"
- "Generate PowerPoint presentations"
- "Batch edit Office documents"

## Domain Knowledge


### Office MCP Tools

| Tool | Input | Output |
|------|-------|--------|
| `create_docx` | Title, sections, styles | .docx file |
| `edit_docx` | Path, changes | Updated .docx |
| `create_xlsx` | Data, formulas | .xlsx file |
| `create_pptx` | Slides, layout | .pptx file |

### Integration with Claude Skills

```markdown
# Example: Combining Skills + MCP

User: "Create a sales report from this data"

1. Data Analysis Skill → Analyze data
2. office-mcp/create_docx → Generate Word report
3. office-mcp/create_xlsx → Generate Excel summary
4. office-mcp/create_pptx → Generate PowerPoint deck
```

### MCP Server Implementation

```python
from mcp import Server
from docx import Document
from openpyxl import Workbook

server = Server("office-mcp")

@server.tool("create_docx")
async def create_docx(title: str, content: str, output_path: str):
    doc = Document()
    doc.add_heading(title, 0)
    doc.add_paragraph(content)
    doc.save(output_path)
    return {"status": "success", "path": output_path}

@server.tool("create_xlsx")
async def create_xlsx(data: list, output_path: str):
    wb = Workbook()
    ws = wb.active
    for row in data:
        ws.append(row)
    wb.save(output_path)
    return {"status": "success", "path": output_path}
```


## Best Practices

1. **Validate inputs before document operations**
2. **Use temp files for large documents**
3. **Return structured responses with file paths**
4. **Handle errors gracefully with meaningful messages**

## Installation

```bash
# Install required dependencies
pip install python-docx openpyxl python-pptx reportlab jinja2
```

## Resources

- [Office MCP Repository](https://github.com/anthropics/skills)
- [Claude Office Skills Hub](https://github.com/claude-office-skills/skills)

Overview

This skill exposes Word, Excel, and PowerPoint operations via an MCP server so an AI agent can create, edit, and assemble Office documents programmatically. It standardizes inputs and outputs for document generation, editing, and batch processing. The goal is repeatable, scriptable Office automation driven by high-level prompts.

How this skill works

The skill registers a set of MCP tools (create_docx, edit_docx, create_xlsx, create_pptx) that accept structured parameters and return file paths and status objects. Each tool builds or modifies files using common Python libraries, writes to an output path (typically a temp or specified directory), and returns a structured response. Tools are intended to be composed with other skills: analysis or data-prep tools feed data into Office tools which then generate final deliverables.

When to use it

  • Generate templated Word reports from AI content or data
  • Produce Excel workbooks with rows, formulas, and formatted sheets
  • Create PowerPoint decks from structured slide outlines
  • Batch-edit existing Office files (replace text, update tables, refresh charts)
  • Automate end-to-end report pipelines combining analysis and document generation

Best practices

  • Validate and sanitize all input data before invoking file operations
  • Use temporary output paths for large or intermediate files, then move to final storage
  • Return explicit, structured responses including status, path, and error details
  • Limit in-memory document sizes; stream or chunk very large datasets
  • Handle exceptions with clear messages so orchestrating agents can retry or fallback

Example use cases

  • Create a weekly sales report: analyze sales data → generate .xlsx summary → produce .docx narrative report → export a .pptx executive summary
  • Batch-update a folder of contracts to apply new header/footer content and version metadata
  • Generate product spec documents from a structured JSON schema into Word and slide highlights into PowerPoint
  • Build financial models in Excel with formulas populated from a data-analysis skill then export charts and tables into a presentation

FAQ

What file formats are supported?

Primary targets are .docx, .xlsx, and .pptx. Outputs are standard Office Open XML files created with python-docx, openpyxl, and python-pptx.

How do I handle large documents or many files?

Write to temporary files, stream data where possible, and paginate content. For bulk operations, process in batches and monitor memory; use disk-backed storage for intermediates.