45 lines
1.1 KiB
Markdown
45 lines
1.1 KiB
Markdown
# docsorter
|
|
|
|
This project provides a JSON-RPC MCP server for working with PowerPoint files.
|
|
|
|
> WILL BE CHANGED TO HEROAGENT WHICH CAN BE CALLED FROM ANYWHERE ALSO AS MCP
|
|
|
|
## Installation
|
|
|
|
To get started, run the installation script and sync the dependencies using `uv`:
|
|
|
|
```bash
|
|
./install.sh
|
|
uv sync
|
|
```
|
|
|
|
## Running the Server
|
|
|
|
You can run the MCP server with the following command:
|
|
|
|
```bash
|
|
uv run docsorter-server
|
|
```
|
|
|
|
## Using the Client
|
|
|
|
The client provides a command-line interface to interact with the server.
|
|
|
|
### Find Presentations
|
|
|
|
```bash
|
|
uv run docsorter-client find --start ~/Documents --pattern "*deck*.pptx"
|
|
```
|
|
|
|
### List Slides
|
|
|
|
```bash
|
|
uv run docsorter-client list-slides --path /path/to/your/presentation.pptx
|
|
```
|
|
|
|
### Notes on Copying Slides
|
|
|
|
The slide copying functionality is implemented in pure Python using `python-pptx` and direct OPC layer manipulation. While this approach supports most common content like text, shapes, and images, it has known limitations for rare or complex embedded content (e.g., OLE objects, videos).
|
|
|
|
If an unsupported content type is detected, the server will return a `CopyUnsupportedError` to ensure predictable behavior.
|