ERP Evolution: The Shift to Agentic Commerce via Model Context Protocol (MCP)
These articles are AI-generated summaries. Please check the original sources for full details.
Economia de Agentes: Tu ERP Necesita Hablar con IAs
McKinsey reports that AI agents will mediate $3 to $5 trillion in global commerce by 2030. This shift is evidenced by the MCP SDK reaching 97 million monthly downloads as of early 2026, a massive jump from 100,000 the previous year.
Why This Matters
While Spanish ERP vendors focus on Verifactu compliance, they risk obsolescence by ignoring the architectural shift toward agentic interaction. Technical reality dictates that without a public API or MCP server, a management system effectively does not exist for the agents that will mediate 90% of B2B purchases by 2028. The division of responsibility is critical: agents provide the business data, but the ERP must provide the regulatory logic and legal compliance which LLMs cannot guarantee independently.
Key Insights
- MCP (Model Context Protocol) has become the industry standard for agent-tool interaction, supported by Anthropic, OpenAI, and Google as of 2025.
- Claude Code reached $2.5 billion in annualized revenue by January 2026, signaling massive developer adoption of agentic coding tools.
- Gartner predicts that 90% of B2B purchases will be mediated by agents by 2028, channeling $15 trillion through automated exchanges.
- Cursor became the fastest SaaS to reach $1 billion in ARR, driven by seamless AI integration into the development workflow in 2025.
- The Agentic AI Foundation, under the Linux Foundation, now governs MCP with platinum members including AWS, Microsoft, and Cloudflare.
Practical Applications
- Automated Invoicing: Using Claude Code to execute ‘create_invoice’ via MCP, ensuring fiscal compliance without manual data entry in the ERP UI.
- Pitfall: Relying on generic LLM knowledge for tax calculation; agents must delegate to ERP logic to ensure legal compliance with Spanish VAT/IRPF rules.
- Contextual Procurement: Developers using Cursor to verify client payment status via ERP API before activating premium features in production.
- Pitfall: Using ‘black box’ ERPs without public APIs, which prevents integration with modern agentic platforms or automation tools like n8n and Zapier.
References:
Continue reading
Next article
GitHub Agentic Workflows: Automating Software Development with Intent-Driven AI
Related Content
Solving the Data Layer Problem in Agentic AI Systems
Production AI agents fail without a structured data layer; the Model Context Protocol (MCP) provides essential real-time ground truth for factual accuracy.
Standardizing AI Tool Integration with the Model Context Protocol (MCP)
Anthropic's Model Context Protocol (MCP) establishes an open standard for AI assistants to call external tools via JSON-RPC, eliminating model-specific function calling fragmentation.
Understanding Model Context Protocol (MCP): A Standardized Bridge for Agentic AI
Anthropic's Model Context Protocol (MCP) standardizes how LLMs securely connect to external data sources, enabling more efficient and scalable agentic workflows across fragmented enterprise APIs.