Model Context Protocol (MCP) and Enterprise Knowledge Retrieval
Integrating muk_mcp, external vector databases, Retrieval-Augmented Generation (RAG), and policy document indexing.
1. Overview & Business Value#
This comprehensive operational guide details the architecture, business workflows, configuration parameters, and execution steps for Model Context Protocol (MCP) and Enterprise Knowledge Retrieval within SmartERP.


Integrating muk_mcp, external vector databases, Retrieval-Augmented Generation (RAG), and policy document indexing.
2. Prerequisites & Security Access Groups#
- Required Modules:
ashewa_ai_workforce, ashewa_ai_bridge, smarterp_ai_agents - Security Groups:
AI Workforce Operator(Technical ID:ashewa_ai.group_ai_operator)AI Systems Administrator(Technical ID:ashewa_ai.group_ai_admin)Governance Officer(Technical ID:ashewa_ai.group_ai_governance)
- Master Data Dependencies: Autonomous AI agent profiles, n8n orchestrator webhooks, LLM model parameters, prompt governance rules, and human-in-the-loop approval queues.
3. Step-by-Step Configuration & Execution#
Step 1: Navigating to the Interface#
- Log into your SmartERP instance at
https://smarterp.et/odoo. - Open the main App Switcher (grid icon in the top-left corner).
- Navigate to AI Workforce → Agents & Workflows.

Step 2: Operational Data Parameters & Controls#
Ensure all mandatory parameters comply with enterprise operational standards:
- Authorized Operator: Action executed under assigned RBAC permissions (AI Operations Architect, Automation Lead, Department Approver).
- Relational Integrity: Linked to upstream master records, approval requests, or transactional ledgers.
- Statutory & Quality Controls: Verified against Ethiopian statutory regulations and corporate governance standards.
Step 3: Operational Execution & State Progression#
- State Progression Gate: Advance record from Draft through Audited to Confirmed / Certified.
- Integrated System Impact: Automatically updates real-time operational states and posts financial/inventory journal moves.
- Audit Verification: System registers tamper-evident audit records logging user ID, timestamp, and cryptographic transaction digests.
The Model Context Protocol (MCP) Standard#
SmartERP implements the open Model Context Protocol (muk_mcp) to enable bi-directional communication between AI agents and external knowledge repositories, enterprise vector stores, and specialized departmental data sources.
Retrieval-Augmented Generation (RAG) Architecture#
- Document Ingestion: Internal enterprise policy manuals, Ethiopian tax proclamations (Proclamation 979/2016, 1156/2019), and ISO engineering standards are ingested into an enterprise vector store.
- Contextual Retrieval: When an AI Copilot receives an inquiry, the MCP server queries semantic vector embeddings to retrieve relevant legal and procedural context clauses.
- Prompt Grounding: The retrieved text is injected into the model prompt context, preventing hallucination and ensuring compliance with official institutional policies.
4. Advanced Configuration Points#
- Automated Actions & Scheduled Tasks (Crons):
AI Bridge: Heartbeat check for n8n orchestrator nodes(ashewa_ai.ir_cron_heartbeat): Runs every 10 minutes to verify gateway connectivityAI Governance: Auto-archive cleared approval queue tickets(ashewa_ai.ir_cron_archive_audit): Runs daily at midnight to maintain governance ledger performance
- Custom Print Templates & Reports:
- AI Agent Operational Execution Audit Trail
- LLM Token Consumption & Cost Allocation Schedule
- Human-in-the-Loop Decision & Override Ledger
- Agent Workflow Error & Latency Analysis
- Multi-Company Rules: Record-level security isolation under Company 1 (
Ashewa Technology Solutions S.C.), preventing cross-company ledger leakage across parent and subsidiary legal entities.
5. Comprehensive Troubleshooting Matrix#
| Error Symptom / Exception | Probable Root Cause | Corrective Action / Resolution |
|---|---|---|
Connection Refused: n8n webhook unreachable | Endpoint URL invalid, orchestrator service offline, or network firewall block. | Check n8n service health and update bridge URL in AI Workforce Settings. |
Token Limit Exceeded: Context window exhausted | Prompt context exceeds configured model maximum token threshold. | Enable automatic context pruning and reduce historical message buffer in agent settings. |
Governance Block: High-risk action requires human approval | Agent attempted financial posting or document deletion above autonomy threshold. | Review and approve the pending transaction in the Human-in-the-Loop Governance Queue. |
Last updated: 2026-11-15