docwriter-mcp-server
A Model Context Protocol server for programmatic creation, modification, and compilation of structured LaTeX documents.
README
docwriter-mcp-server 📄✍️
A Model Context Protocol (MCP) server for programmatic creation, modification, and compilation of structured LaTeX documents.
This server provides a suite of tools for an AI agent or other MCP client to manage the lifecycle of a document on the local filesystem, from bootstrapping from a template to applying structured updates and compiling the final PDF output. It is built on the robust cyanheads/mcp-ts-template.
This project is in beta - some things may be broken. Please report any issues or feature requests on GitHub.
🚀 Core Capabilities: Document Tools 🛠️
This server equips your AI with specialized tools to create and manage LaTeX documents:
| Tool Name | Description | Key Features |
|---|---|---|
docwriter_create_latex_document |
Creates a new .tex file from a template. |
- Bootstrap from simple_report, ieee_article, or research_report templates.<br/>- Populates title and author metadata. |
docwriter_update_document_block |
Updates one or more named content blocks within a document. | - Atomically updates multiple sections (e.g., abstract, introduction).<br/>- Preserves document structure.<br/>- Securely sanitizes all input content. |
docwriter_search_and_replace |
Performs a simple, global search and replace for text. | - Useful for quick, non-structural text changes.<br/>- Sanitizes replacement text. |
docwriter_compile_latex_to_pdf |
Compiles a .tex document into a PDF. |
- Uses lualatex with multiple passes to resolve cross-references.<br/>- Automatically runs biber for bibliography processing.<br/>- Returns compilation logs for debugging. |
docwriter_list_latex_documents |
Retrieves a list of all available documents. | - Scans the data directory for all .tex files. |
Table of Contents
| Overview | Features | Installation |
|---|---|---|
| Configuration | Project Structure | Development |
| License |
Overview
The docwriter-mcp-server acts as a specialized backend, allowing MCP-compatible clients—such as AI agents, IDE extensions, or automated workflows—to programmatically generate and manage professional-quality LaTeX documents.
Instead of manual document preparation, your tools can leverage this server to:
- Automate Report Generation: Create consistent, templated reports, articles, or papers.
- Dynamically Insert Content: Populate documents with data, analysis, or text generated by an AI.
- Ensure Document Quality: Compile and verify documents as part of a CI/CD pipeline.
- Integrate with AI Workflows: Enable LLMs to create and edit complex documents as part of a larger task.
Developer Note: This repository includes a .clinerules file that serves as a developer cheat sheet for your LLM coding agent with quick reference for the codebase patterns, file locations, and code snippets.
Features
Core Utilities
Leverages the robust utilities provided by the mcp-ts-template:
- Logging: Structured, configurable logging with sensitive data redaction.
- Error Handling: Centralized error processing and standardized error types (
McpError). - Configuration: Type-safe environment variable loading with Zod validation.
- Input Sanitization: Strong security focus with utilities for sanitizing LaTeX, HTML, and file paths.
- HTTP Transport: High-performance HTTP server using Hono, featuring session management, CORS, and rate limiting.
- Authentication: Robust authentication layer supporting JWT and OAuth 2.1.
Document Generation
- Template-Based Creation: Start documents from
simple_report,ieee_article, orresearch_reporttemplates. - Structured Updates: Safely modify content within
%% -- BLOCK: ... -- %%markers. - Secure Compilation: Executes
lualatexwith multiple passes, automatically runningbiberfor bibliographies and cleaning up auxiliary files. - Filesystem Backend: All documents and outputs are stored and managed on the local filesystem in a configurable data directory.
Installation
Prerequisites
- Node.js (>=20.0.0)
- npm (comes with Node.js)
- A full TeX Live distribution (or equivalent like MiKTeX). The
lualatexandbibercommands must be in the system'sPATH.
Install from Source
-
Clone the repository:
git clone https://github.com/cyanheads/docwriter-mcp-server.git cd docwriter-mcp-server -
Install dependencies:
npm install -
Build the project:
npm run build
Configuration
Configure the server using environment variables in a .env file.
| Variable | Description | Default |
|---|---|---|
DOCWRITER_DATA_PATH |
Required. The root directory for storing .tex files and compiled PDFs. |
./data |
MCP_TRANSPORT_TYPE |
Server transport: stdio or http. |
stdio |
MCP_LOG_LEVEL |
Logging level (debug, info, warning, error). |
debug |
MCP_HTTP_PORT |
Port for the HTTP server. | 3010 |
MCP_AUTH_MODE |
Authentication mode for HTTP: jwt or oauth. |
jwt |
MCP_AUTH_SECRET_KEY |
Required for jwt mode. Secret key (min. 32 chars) for signing tokens. |
(none) |
OAUTH_ISSUER_URL |
Required for oauth mode. The issuer URL of your OAuth 2.1 provider. |
(none) |
OAUTH_AUDIENCE |
Required for oauth mode. The audience identifier for this server. |
(none) |
Project Structure
The codebase follows a modular structure within the src/ directory:
src/
├── index.ts # Entry point: Initializes and starts the server
├── config/ # Configuration loading (env vars, package info)
│ └── index.ts
├── mcp-server/ # Core MCP server logic and capability registration
│ ├── server.ts # Server setup, tool registration
│ ├── transports/ # Transport handling (stdio, http)
│ └── tools/ # MCP Tool implementations (subdirs per tool)
├── types-global/ # Shared TypeScript type definitions
└── utils/ # Common utility functions (logger, error handler, etc.)
For a detailed file tree, run npm run tree.
Development
Build and Test
# Build the project (compile TS to JS and make executable)
npm run build
# Test the server locally using the MCP inspector tool (stdio transport)
npm run inspector
# Clean build artifacts
npm run clean
# Clean build artifacts and then rebuild the project
npm run rebuild
# Format code with Prettier
npm run format
# Start the server using stdio (default)
npm start
# Start the server using HTTP transport
npm run start:http
License
This project is licensed under the Apache License 2.0 - see the LICENSE file for details.
<div align="center"> Built with ❤️ and the <a href="https://modelcontextprotocol.io/">Model Context Protocol</a> </div>
推荐服务器
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。
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。
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。
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。
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。
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。
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。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
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。