@docujson/mcp
Connects AI assistants to DocuJSON for PDF generation, template management, and API operations via natural language. It enables tasks like creating invoices, managing templates, and viewing usage stats through MCP-compatible tools.
README
@docujson/mcp
DocuJSON Model Context Protocol (MCP) server. Connect AI assistants like Claude Desktop, Cursor, and other MCP-compatible tools to DocuJSON for PDF generation, template management, and API key operations via natural language.
What it does
This package is a thin stdio bridge that forwards MCP requests from your AI assistant to the hosted DocuJSON MCP server at https://mcp.docujson.com.
With this connected, you can ask your AI assistant things like:
- "Generate an invoice PDF for Acme Corp with 3 line items totaling $4,500"
- "Create a weekly status report template"
- "Show me my PDF usage this month"
- "List all my custom templates"
- "Revise the invoice template to use a blue header"
Installation
You don't install it directly — your MCP client (Claude Desktop) runs npx @docujson/mcp on demand.
Setup
1. Get a DocuJSON API key
- Sign up at docujson.com/signup
- Go to Dashboard → API Keys
- Create a new key (starts with
dj_)
2. Configure your MCP client
Claude Desktop
Edit your Claude Desktop config file:
- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\Claude\claude_desktop_config.json
Add the DocuJSON server:
{
"mcpServers": {
"docujson": {
"command": "npx",
"args": ["-y", "@docujson/mcp"],
"env": {
"DOCUJSON_API_KEY": "dj_your_api_key_here"
}
}
}
}
Restart Claude Desktop. You should see DocuJSON tools available in the tools panel.
Cursor (HTTP — recommended)
Cursor supports the HTTP transport directly, so you don't need this stdio bridge. Use the hosted server URL:
Cursor Settings → MCP → Add server:
- Name:
docujson - URL:
https://mcp.docujson.com?apiKey=dj_your_api_key
Continue, Zed, or other MCP clients
Use the HTTP URL if your client supports it. Otherwise use the stdio config pattern from the Claude Desktop example above.
Available tools
generate_pdf— Generate a PDF from any templatepreview_template— Generate a preview (doesn't count toward quota)list_templates— List all built-in and custom templatesget_template— Get template details and JSON schemacreate_custom_template— AI-generate a new template from a descriptionupdate_custom_template— Update template metadata or sample datadelete_custom_template— Delete a custom template (requires confirmation)revise_template— Submit a natural-language revision requestpublish_template— Publish a draft templatelist_api_keys— List workspace API keyscreate_api_key— Create a new API keyrevoke_api_key— Revoke (deactivate) an API keyget_usage— Monthly PDF usage statsget_workspace— Current workspace informationlist_integrations— Supported integration platformsget_integration_script— Code snippet for a specific integration
Full tool reference: docujson.com/mcp
Environment variables
DOCUJSON_API_KEY(required) — Your DocuJSON API keyDOCUJSON_MCP_URL(optional) — Override the MCP server URL. Defaults tohttps://mcp.docujson.com
Troubleshooting
"Missing DOCUJSON_API_KEY"
Set the environment variable in your MCP client config (see Setup above).
Tools don't show up in Claude Desktop
- Restart Claude Desktop completely
- Check the developer logs: View → Developer → Open Developer Tools
- Verify the config JSON is valid (trailing commas break it)
"Unauthorized" errors
- Verify your API key is active in the DocuJSON dashboard
- Make sure it starts with
dj_
License
MIT
Links
- Documentation: docujson.com/mcp
- API Reference: docujson.com/docs
- Dashboard: docujson.com/dashboard
- Support: support@docujson.com
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。