Elorus MCP Server
Read-only MCP server for the Elorus invoicing platform, enabling natural language queries about invoices, contacts, expenses, payments, and products.
README
Elorus MCP Server
A read-only Model Context Protocol server for the Elorus invoicing platform — ask your business data questions in plain language.
Unofficial & read-only. This project is not affiliated with Elorus. It only ever issues GET requests — it cannot create, modify, or delete anything in your organization.
What can you ask?
- "Which invoices are overdue, and for how much?"
- "How much did we invoice ACME Ltd in July?"
- "What did we spend on hosting last month?"
- "Which payments did we receive this week?"
- "What is the VAT number of client X?"
Features
- 10 read-only tools covering invoices, contacts, expenses, cash receipts and products (Elorus API v1.2)
- Curated filters — period, status, paid/unpaid, overdue, client/supplier, free-text search — documented in every tool schema so the LLM calls them correctly
- Token-efficient responses — list tools return trimmed summaries;
get_*tools return the full record - Read-only by construction — the HTTP client implements GET only
- Clear error messages for authentication, rate-limit and connectivity problems
Architecture
flowchart LR
A["MCP client (Claude Desktop / Claude Code)"] -- stdio --> B["elorus-mcp (MCP Python SDK)"]
B --> C["ElorusClient (GET-only, httpx)"]
C -- "Authorization: Token + X-Elorus-Organization" --> D["Elorus API v1.2"]
Quickstart
1. Get your credentials
- API key: Elorus web app → User Profile → API token
- Organization ID: Elorus web app → Settings → Organization → Organization ID
2. Configure your MCP client
Claude Desktop (claude_desktop_config.json):
{
"mcpServers": {
"elorus": {
"command": "uvx",
"args": ["--from", "git+https://github.com/nikagabriel741agent/Elorus-ERP-MCP-Server", "elorus-mcp"],
"env": {
"ELORUS_API_KEY": "your-api-key",
"ELORUS_ORGANIZATION_ID": "your-organization-id"
}
}
}
}
Claude Code:
claude mcp add elorus \
-e ELORUS_API_KEY=your-api-key \
-e ELORUS_ORGANIZATION_ID=your-organization-id \
-- uvx --from git+https://github.com/nikagabriel741agent/Elorus-ERP-MCP-Server elorus-mcp
Local development install
git clone https://github.com/nikagabriel741agent/Elorus-ERP-MCP-Server.git
cd Elorus-ERP-MCP-Server
uv sync
cp .env.example .env # fill in your credentials
uv run --env-file .env elorus-mcp
Tools
| Tool | Description | Key arguments |
|---|---|---|
list_invoices |
List sales invoices (summaries) | period_from, period_to, status, paid, overdue, draft, client, search, ordering, page, page_size |
get_invoice |
Full invoice record | invoice_id |
list_contacts |
List clients & suppliers | search, contact_type (client/supplier), active, page, page_size |
get_contact |
Full contact record | contact_id |
list_expenses |
List expenses | period_from, period_to, supplier, search, page, page_size |
get_expense |
Full expense record | expense_id |
list_cash_receipts |
List payments received | period_from, period_to, contact, invoice, transaction_type, page, page_size |
get_cash_receipt |
Full cash receipt record | cash_receipt_id |
list_products |
List products & services | search, active, sales, purchases, page, page_size |
get_product |
Full product record | product_id |
Notes: dates are YYYY-MM-DD; period_from/period_to only apply when both are set.
Invoice status is one of draft, pending, issued, partial, unpaid, overdue, paid, void.
Configuration
| Environment variable | Required | Description |
|---|---|---|
ELORUS_API_KEY |
yes | Personal API key (Elorus web app → User Profile) |
ELORUS_ORGANIZATION_ID |
yes | Elorus web app → Settings → Organization → Organization ID |
ELORUS_BASE_URL |
no | Defaults to https://api.elorus.com/v1.2 |
Development
uv sync # install dependencies
uv run pytest -v # run the test suite (no real API calls)
uv run ruff check . # lint
uv run ruff format --check . # formatting
License
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。