onec-odata-mcp
Bridges AI agents to 1C:Enterprise via OData, enabling natural language queries of business entities like counterparties and invoices with read-only access.
README
onec-odata-mcp
MCP server bridging AI agents to 1C:Enterprise (1С:Предприятие) via OData.
Give an AI assistant (Claude Desktop, Cursor, any MCP client) read-only access to a 1C database. Instead of exporting spreadsheets or hand-writing OData URLs, the agent queries 1C entities — counterparties, documents, invoices, catalogs — in natural language and gets structured data back. Works with any 1C configuration that exposes the standard OData endpoint (Accounting, ERP, Trade Management, and custom configs).
Quick start
npm install
npm run build
Configure via environment variables (recommended for MCP clients):
ONEC_BASE_URL=https://1c.example.com/your_db/odata/standard.odata
ONEC_USERNAME=odata_user
ONEC_PASSWORD=odata_password
# optional:
ONEC_DATABASE=your_db
ONEC_METADATA_CACHE_TTL_MS=3600000
Register the server in your MCP client (e.g. Claude Desktop claude_desktop_config.json):
{
"mcpServers": {
"onec-odata": {
"command": "npx",
"args": ["-y", "onec-odata-mcp"],
"env": {
"ONEC_BASE_URL": "https://1c.example.com/your_db/odata/standard.odata",
"ONEC_USERNAME": "odata_user",
"ONEC_PASSWORD": "your_password"
}
}
}
}
Alternatively, place a JSON file at ~/.onec-odata/onec-config.json with { "baseUrl", "username", "password" } and point ONEC_CONFIG_PATH at it.
Tools
| Tool | Description |
|---|---|
odata_config |
Check configuration status and setup instructions |
odata_list_entities |
List all available OData entities |
odata_metadata |
Get entity types and sets from $metadata (cached, 1h TTL) |
odata_explain_entity |
Describe an entity's fields, types, keys, navigation properties |
odata_build_query |
Build a validated $filter/$select/$orderby from structured intent |
odata_query |
Query an entity with raw OData params (filter/select/top/expand) |
odata_count |
Count records in an entity, optionally filtered |
odata_financial_summary |
Auto-detect common financial entities and report their counts |
Example
Ask the agent: "How many sales orders were posted this month?"
It calls odata_build_query → odata_count on the sales-order entity with a date filter. No OData syntax knowledge required — the server validates fields against live $metadata and rejects unknown/secret fields.
Requirements
- Node.js ≥ 18
- 1C:Enterprise with the OData endpoint published (web publishing →
standard.odata)
License
MIT
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。