Aimino-Tech/opendocswork-mcp
FreeRust 原生、本地优先的 Office 文档处理 MCP 服务器,支持 Excel、Word、PowerPoint,亚毫秒级响应。
About Aimino-Tech/opendocswork-mcp
opendocswork-mcp (also known as office-oxide-mcp) is a Rust-native MCP (Model Context Protocol) server for high-performance Office document processing. It supports Excel (XLSX), Word (DOCX), PowerPoint (PPTX), and PDF formats with sub-millisecond latency. The server is local-first and open source, designed to treat PDFs as editable forms — filling AcroForms, XFA forms, and even flat scanned PDFs using coordinate-based input. It integrates seamlessly with AI assistants like Claude Desktop, Cursor, and VS Code Copilot via MCP. The tool can read documents into JSON, Markdown, Chunks, or plain text; generate business reports, invoices, presentations, and financial statements; list supported formats; retrieve document metadata; and export to PDF. It aims to be the 'open source Aspose' for Office automation.
Key Features
Pros & Cons
- Local-first: no data sent to external servers, ensuring privacy and security
- Sub-millisecond latency enables real-time document processing in AI workflows
- Open source with permissive license, allowing customization and community contributions
- Supports a wide range of Office and PDF formats natively
- Handles complex PDF forms, including XFA and scanned documents with coordinate mapping
- Seamless integration with popular AI assistants via standard MCP protocol
- Rust implementation provides high performance and memory safety
- No dependency on cloud APIs or third-party services for document processing
- Installation requires Rust toolchain if using cargo install, which may be unfamiliar for non-developers
- Primarily designed for local use; no built-in cloud deployment options
- Documentation is currently limited to the GitHub README and examples in the showcase folder
- Requires an MCP-compatible client (e.g., Claude Desktop, Cursor) to leverage full capabilities
- May have a learning curve for users unfamiliar with MCP servers and JSON-RPC calls
- Actively in early development; features and stability may change