Holded MCP Server
An open-source MCP server for the Holded API that enables natural language management of invoicing, CRM, projects, team, and accounting.
README
Holded MCP Server
An open-source Model Context Protocol (MCP) server for the Holded API. Connect any MCP-compatible AI assistant to your Holded account and manage invoicing, CRM, projects, team, and accounting through natural language.
Built with Python, FastMCP, and httpx.
Features
- Invoicing — Create and manage invoices, estimates, sales orders, credit notes, receipts, proformas, and purchase documents.
- Contacts — Full CRUD for clients, suppliers, leads, debtors, and creditors.
- Products — Manage your product catalog, pricing, stock levels across warehouses.
- CRM — Work with sales funnels, leads/deals, and activities.
- Projects — Create projects, tasks, and track time.
- Team / HR — Manage employees, clock in/out, and attendance records.
- Accounting — Access the chart of accounts and daily ledger entries.
- Treasury — List and create bank/cash accounts.
Setup
Prerequisites
- Python 3.10+
- A Holded account with an API key (Settings > API)
Installation
pip install -e .
# Set your Holded API key
export HOLDED_API_KEY=your_api_key_here
Configuration
HOLDED_API_KEY (required)
Your Holded API key. You can find it in Holded under Settings > API.
HOLDED_ALLOWED_METHODS (optional)
Controls which HTTP methods the server is permitted to use. This acts as a safety mechanism to prevent unintended write or delete operations.
| Value | Behavior |
|---|---|
ALL (default) |
All methods allowed — full read/write/delete access |
GET |
Read-only mode — only list and get operations work |
GET,POST |
Read + create — no updates or deletions |
GET,POST,PUT |
Read, create, and update — no deletions |
GET,POST,PUT,DELETE |
Same as ALL — explicit full access |
Examples:
# Read-only mode (safe for exploration)
export HOLDED_ALLOWED_METHODS=GET
# Allow reads and creating new records
export HOLDED_ALLOWED_METHODS=GET,POST
# Full access (default)
export HOLDED_ALLOWED_METHODS=ALL
Tip: Start with
GETto safely explore your data, then expand permissions as needed.
Usage
Claude Code
claude mcp add holded -- python -m holded_mcp.server
Or add to .claude/settings.json:
{
"mcpServers": {
"holded": {
"command": "python",
"args": ["-m", "holded_mcp.server"],
"env": {
"HOLDED_API_KEY": "your_api_key_here"
}
}
}
}
Cursor
Add to your Cursor MCP configuration (.cursor/mcp.json):
{
"mcpServers": {
"holded": {
"command": "python",
"args": ["-m", "holded_mcp.server"],
"env": {
"HOLDED_API_KEY": "your_api_key_here"
}
}
}
}
Windsurf
Add to your Windsurf MCP configuration (~/.windsurf/mcp.json):
{
"mcpServers": {
"holded": {
"command": "python",
"args": ["-m", "holded_mcp.server"],
"env": {
"HOLDED_API_KEY": "your_api_key_here"
}
}
}
}
VS Code (Copilot)
Add to your VS Code settings (.vscode/mcp.json):
{
"servers": {
"holded": {
"command": "python",
"args": ["-m", "holded_mcp.server"],
"env": {
"HOLDED_API_KEY": "your_api_key_here"
}
}
}
}
Any MCP-Compatible Client
The server uses stdio transport by default. Run it with any MCP client that supports stdio:
python -m holded_mcp.server
Or use the installed entry point:
holded-mcp
Available Tools
Contacts
list_contacts, get_contact, create_contact, update_contact, delete_contact
Documents
list_documents, get_document, create_document, update_document, delete_document, pay_document, send_document, get_document_pdf
Document types: invoice, salesreceipt, creditnote, estimate, salesorder, waybill, proform, purchase, purchaserefund, purchaseorder
Products
list_products, get_product, create_product, update_product, delete_product, update_stock
Treasury
list_treasuries, create_treasury
CRM
list_funnels, list_leads, get_lead, create_lead, update_lead, delete_lead, list_events, create_event
Projects
list_projects, get_project, create_project, update_project, delete_project, list_tasks, create_task, update_task, list_time_records, create_time_record
Team
list_employees, get_employee, create_employee, update_employee, clock_in, clock_out, list_time_entries
Accounting
list_daily_ledger, create_ledger_entry, list_accounts, get_account, create_account
Use Cases
- Automate invoicing — Ask your AI assistant to create invoices, send them to clients, and track payments.
- CRM management — Manage your sales pipeline, create leads, and log activities without leaving your editor.
- Project tracking — Create projects and tasks, log time, and monitor progress through conversation.
- HR operations — Manage employees, track attendance, and handle clock in/out.
- Financial overview — Query your chart of accounts, review ledger entries, and manage treasury accounts.
- Bulk operations — Use AI to process multiple records, generate reports, or migrate data.
Architecture
The server follows a modular architecture:
- Entry point (
server.py) — Creates the FastMCP instance and registers all tool modules. - API client (
client.py) — Async HTTP client with auth, method restrictions, and pagination support. - Tool modules (
tools/*.py) — Each module exports aregister(mcp, client)function. Modules are purely functional with no cross-dependencies.
Author
Built by Javier Chulvi.
License
MIT
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.