sap-sd-mcp-validation

sap-sd-mcp-validation

Validates SAP SD sales orders (material, pricing, credit) before creation using the MCP protocol, enabling clean-core order-to-cash automation.

Category
Visit Server

README

🤖 SAP SD Order Validation Agent — MCP Protocol

Open In Colab Python MCP SAP Clean Core

Autor: Fabian Andrés Acevedo R. | Bogotá, Colombia Certificaciones: C_S4CPB_2602 · C_LCNC_2601 LinkedIn: fabian-acevedo-sap-cloud-data


🎯 ¿Qué hace este proyecto?

Un agente de inteligencia artificial que valida órdenes de venta SAP SD (Order-to-Cash) antes de impactar el ERP, utilizando el Model Context Protocol (MCP) de Anthropic como capa de comunicación entre Claude/Joule y SAP S/4HANA Cloud Public Edition.

El problema que resuelve

En SAP SD, los errores de pricing, crédito y disponibilidad de material generan correcciones manuales costosas:

Métrica Sin agente Con agente MCP
Error rate en órdenes ~3-5% < 0.5%
Tiempo de corrección/error 30-45 min 0 min
Costo estimado/error ~$400 USD $0
Ahorro mensual (500 órdenes) ~$5,400 USD

🏗️ Arquitectura Clean Core

Usuario / Comprador
        ↓
   MCP Client
(Our Server / App)
        ↓
  SAP SD MCP Server         ← Este repositorio
  (sap_sd_server.py)
        ↓                         ↓
  Claude / Joule         S/4HANA Cloud API
     Agent               (API_SALES_ORDER_SRV)

Principio Clean Core: Toda la lógica de validación corre en la periferia del ERP a través de MCP. El core de SAP S/4HANA no es modificado — extensibilidad 100% vía BTP.


🛠️ Tools expuestas al agente

Tool Descripción Equivalente SAP
validate_material_number Verifica existencia y stock Verificación ATP
check_pricing_conditions Calcula precio total de la orden Esquema de cálculo
verify_customer_credit_limit Valida cupo financiero del cliente Control de crédito FI
create_sales_order Crea la orden si todo es válido VA01 / API_SALES_ORDER_SRV

Resources disponibles

  • sap://materials — Catálogo de materiales
  • sap://customers — Maestro de clientes
  • sap://orders — Órdenes creadas en sesión

Prompt pre-configurado

  • validate_full_order — Orquesta el flujo completo O2C con auditoría

🚀 Cómo probarlo

Opción A — Google Colab (recomendado, sin instalación)

  1. Haz clic en el badge Open In Colab al inicio de este README
  2. Ve al panel izquierdo → ícono 🔑 Secrets → agrega ANTHROPIC_API_KEY
  3. Menú → Ejecutar todo

Opción B — Local

# Clonar repositorio
git clone https://github.com/TU_USUARIO/sap-sd-mcp-validation.git
cd sap-sd-mcp-validation

# Crear entorno virtual
python -m venv .venv
source .venv/bin/activate  # Windows: .venv\Scripts\activate

# Instalar dependencias
pip install -r requirements.txt

# Configurar API key
echo "ANTHROPIC_API_KEY=sk-ant-tu-key" > .env

# Correr el servidor MCP directamente
python sap_sd_server.py

🧪 Casos de prueba incluidos

Escenario Cliente Material Cantidad Resultado
✅ Happy path CUST-500 MAT-1001 5 Orden SO-1001 creada
❌ Sin stock CUST-500 MAT-1003 1 Falla: stock = 0
❌ Crédito insuficiente CUST-501 MAT-1001 10 Escala a Build Process Automation
❌ Material inválido CUST-500 MAT-9999 1 Falla: material no existe

📋 Audit Log integrado

Cada llamada del agente queda registrada con trazabilidad completa:

[1] 2026-06-18T10:30:01
    Evento : TOOL_CALL
    Tool   : validate_material_number
    Input  : {"material_id": "MAT-1001"}
    Output : {"stock_disponible": 42, "disponible": true}

[2] 2026-06-18T10:30:02
    Evento : TOOL_CALL
    Tool   : check_pricing_conditions
    Input  : {"material_id": "MAT-1001", "quantity": 5}
    Output : {"total_price_usd": 6250.0}

🗺️ Roadmap

Versión Feature Stack adicional
V1 ✅ Material · Pricing · Crédito · Creación FastMCP + Mock data
V2 check_atp() — Available-to-Promise con fechas de entrega Scheduling logic
V3 propose_delivery_split() — Propuesta de entrega parcial Supply Chain logic
V4 trigger_credit_approval() — Escalación vía workflow SAP Build Process Automation
V5 APIs reales S/4HANA Cloud via Communication Arrangement httpx + OAuth2 + BTP Destination

🔗 Alineación con SAP Roadmap oficial

Este proyecto anticipa funcionalidades del SAP S/4HANA Cloud release 2608 (Q3 2026):

  • AI-assisted sales order creation from unstructured data — SAP Road Map Explorer
  • Autonomous Enterprise agents for O2C processes — SAP Sapphire 2026

📁 Estructura del repositorio

sap-sd-mcp-validation/
│
├── .gitignore                          # Excluye .env, __pycache__, .venv
├── requirements.txt                    # Dependencias Python
├── sap_sd_server.py                    # Servidor MCP SAP SD (FastMCP)
├── SAP_SD_MCP_Agent_Fabian_Acevedo.ipynb  # Notebook Colab interactivo
└── README.md                           # Este archivo

🤝 Sobre el autor

SAP Certified Implementation Consultant (C_S4CPB_2602) y SAP Certified Build Developer (C_LCNC_2601). Especializado en SAP BTP, Clean Core extensibility, y automatización de procesos O2C con IA.

Open to remote SAP consulting roles — LATAM / Global (EST/CST)

Recommended Servers

playwright-mcp

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.

Official
Featured
TypeScript
Magic Component Platform (MCP)

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.

Official
Featured
Local
TypeScript
Audiense Insights MCP Server

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.

Official
Featured
Local
TypeScript
VeyraX MCP

VeyraX MCP

Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.

Official
Featured
Local
graphlit-mcp-server

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.

Official
Featured
TypeScript
Kagi MCP Server

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.

Official
Featured
Python
E2B

E2B

Using MCP to run code via e2b.

Official
Featured
Neon Database

Neon Database

MCP server for interacting with Neon Management API and databases

Official
Featured
Exa Search

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.

Official
Featured
Qdrant Server

Qdrant Server

This repository is an example of how to create a MCP server for Qdrant, a vector search engine.

Official
Featured