PPTX

You run a dedicated MCP server that lets AI assistants create, read, and edit PowerPoint presentations programmatically. By exposing a Python-based MCP endpoint you can drive .pptx files from natural language requests, generate thumbnails, and manipulate slides and content with structured commands. This enables seamless integration of presentation workflows into AI copilots and automated pipelines.

How to use

You start the server locally and connect an MCP client to it. The server accepts requests to create new presentations, extract and replace text, rearrange slides, generate thumbnail previews, and unpack or rebuild Office documents. Use it to drive PowerPoint workflows from AI prompts by sending structured instructions that map to the available tools.

How to install

Prerequisites: Python 3.10 or newer is required.

Clone the project, install in editable mode, and start the MCP server.

git clone https://github.com/YOUR_USERNAME/pptx-mcp-server.git
cd pptx-mcp-server

# Install the package in editable mode
pip install -e .

Configure your MCP client to connect to the server using the stdio command shown below. Create or edit your MCP client configuration to include a server named pptx with the following command and arguments.

{
  "mcpServers": {
    "pptx": {
      "command": "python",
      "args": ["-m", "pptx_mcp_server"]
    }
  }
}

Restart your MCP client to load the new server configuration.

Configuration and runtime details

The MCP server is run locally via Python. Use the following exact command to start the server in a development or local environment.

python -m pptx_mcp_server

If you are using a virtual environment, provide the full path to your Python interpreter in the client configuration instead of a generic python command.

Usage examples

Create a new presentation from scratch is one of the core capabilities. You can specify slide layouts, backgrounds, shapes, and text with formatting. The server supports extracting text with positions, replacing text, rearranging slides, and generating thumbnail grids for quick previews.

Common actions you can perform include creating presentations, extracting text inventories, applying text replacements, rearranging slides, generating thumbnails, and packing/unpacking Office documents for XML editing.

Troubleshooting and notes

If the MCP server does not appear in your client, ensure the config is loaded by your client and that you restarted the client after changes. Check that Python is the correct interpreter for your environment and that dependencies are installed.

Appendix: Tool capabilities

The server includes the following operations: create_presentation, extract_text_inventory, apply_text_replacements, rearrange_slides, create_thumbnail_grid, unpack_office_document, pack_office_document, and validate_office_document. These allow you to build end-to-end presentation workflows entirely through structured requests.

Available tools

create_presentation

Create new PowerPoint presentations from scratch with configurable slides, layouts, backgrounds, and shapes.

extract_text_inventory

Retrieve all text content with positional and formatting data from an existing presentation.

apply_text_replacements

Replace text in a presentation using a JSON specification to map old text to new values.

rearrange_slides

Duplicate, delete, and reorder slides within a presentation.

create_thumbnail_grid

Generate a visual grid of slide thumbnails for quick previews.

unpack_office_document

Extract Office files to editable XML for advanced customization.

pack_office_document

Rebuild Office documents from edited XML back into a PPTX.

validate_office_document

Validate the integrity and structure of the Office document.