Frihet Launches Spain's First Official Open-Source MCP Server for ERP
These articles are AI-generated summaries. Please check the original sources for full details.
Frihet lanza el primer servidor MCP oficial de un ERP español
Frihet has launched the first official Model Context Protocol (MCP) server for a Spanish ERP. The system exposes 31 tools across 6 resources, allowing AI assistants like Claude to manage financial data directly without manual data entry.
Why This Matters
Traditional ERP integration requires custom code and significant development investment to bridge the gap between AI assistants and business data. By implementing MCP, Frihet replaces high-friction REST API calls with a standardized conversational protocol, allowing AI models to execute business logic directly. This transition from manual API management to standardized tool-calling reduces the barrier for non-developers to automate complex financial workflows. It shifts the paradigm from mechanical automation to intelligent assistance where the software understands natural language intent.
Key Insights
- Frihet released 31 specific tools for ERP management in 2026 including invoice and expense modules.
- Model Context Protocol (MCP) was established by Anthropic as a universal standard for AI tool integration.
- The @frihet/mcp-server package enables local stdio communication for development tools like Cursor and Windsurf.
- Cloudflare Workers power the remote mcp.frihet.io implementation to support distributed teams.
- The server is released under the MIT license to ensure transparency in processing sensitive financial data.
Working Examples
Command to install and run the Frihet MCP server locally via stdio.
npx @frihet/mcp-server
Practical Applications
- Freelance designers using Claude to generate tax-compliant invoices through natural language commands. Pitfall: Missing FRIHET_API_KEY environment variables will cause authentication errors.
- Development teams querying real-time product sales data within Cursor to maintain coding context. Pitfall: Incorrect stdio configuration in IDE settings prevents the AI from discovering available tools.
References:
Continue reading
Next article
Guía Completa de Facturación para Autónomos en España (2026)
Related Content
Grounding LLMs in Maritime Data: Using MCP for Port Intelligence
Leveraging the Model Context Protocol (MCP) to generate port briefings using real-time data from 16 VesselAPI maritime tools.
Google Colab MCP Server: Programmatic AI Agent Access to GPU Cloud Runtimes
Google releases the open-source Colab MCP Server, enabling AI agents to autonomously execute Python code and manage cloud-hosted GPU runtimes via the Model Context Protocol.
Martina Zrnec Launches Stacky: Bridging Content Hubs and AI Assistants via MCP
Martina Zrnec releases Stacky, an MCP server enabling AI assistants to manage content hubs for 14,000 users on the Stacklist platform.