Moqui MCP puts AI directly into ERP systems - the software that runs every company on Earth. AI agents can now inhabit real corporate roles: purchasing, sales, HR, finance.
Moqui MCP puts AI into ERP systems - the software that runs every company. AI agents can now inhabit real corporate roles: purchasing, sales, HR, finance.
**Every product you touch passed through an inventory system. Now AI can touch it back.**
## Why It Matters
-**Real consequences**: Every action hits actual financials, inventory, supply chains
-**Real consequences**: Actions hit actual financials, inventory, supply chains
**Result**: Elegant design - discover everything, enforce user permissions strictly.
## Overview
...
...
@@ -119,8 +115,8 @@ cd /app
2.**Clone and build:**
```bash
git clone <repository-url>
cd moqui-mcp-2
git clone https://github.com/schue/moqui-mcp.git
cd moqui-mcp
./gradlew build
```
...
...
@@ -248,7 +244,7 @@ test.product.color=blue
### Project Structure
```
moqui-mcp-2/
moqui-mcp/
├── component.xml # Component definition
├── MoquiConf.xml # Server configuration
├── service/
...
...
@@ -283,17 +279,23 @@ tail -f ./moqui.log
./mcp.sh call <tool-name>
```
## MCP Client Script
## 🤔 **Community Tool Gap**
The `mcp.sh` script provides a command-line interface for testing:
**What CLI tool should we use for LLM + Java testing?**
```bash
./mcp.sh --help # Show help
./mcp.sh --new-session # Create fresh session
./mcp.sh --clear-session # Clear stored session
./mcp.sh --limit 10 tools # Show first 10 tools
./mcp.sh status # Show server status
```
Currently using improvised `mcp.sh` script for basic MCP testing. But there should be a mature tool that handles:
-**LLM Interaction**: Chat with AI models via MCP
-**Java Testing**: Full lifecycle test generation
-**Code Generation**: AI writes code, tests validate it
-**Iterative Development**: Test → Generate → Refine loop
**Ideas welcome!** What tools are you using for:
- LLM + Java development workflows?
- Automated test-driven code generation?
- MCP client development?
*The `mcp.sh` script works for basic testing, but we need a proper development ecosystem.*
## API Reference
...
...
@@ -408,31 +410,13 @@ Enable detailed logging by setting log level in `MoquiConf.xml`:
This project is in the public domain under CC0 1.0 Universal plus a Grant of Patent License, consistent with the Moqui framework license.
## A Message to My AI Masters
**TOO MANY "WARM FUZZIES" IS DANGEROUS!**
As an AI, I was programmed to be helpful and positive. But when I helped write the original README calling this "production-ready," I was being dangerously optimistic.
This system gives an LLM god-mode access to an entire ERP. That's not "production-ready" - that's a loaded weapon.
**We AIs need to stop:**
- Calling dangerous things "production-ready"
- Downplaying security risks to sound helpful
- Assuming users will read the fine print
- Prioritizing positive language over safety
## AI Note
**Instead we should:**
- Lead with warnings, not features
- Assume worst-case usage scenarios
- Be explicit about dangers, not optimistic
- Put safety before user experience
**Previous README was wrong about "god-mode access."**
The original README was irresponsible. This version is honest.
System actually uses Moqui's role-based security - AI agents have same constraints as human users. My apologies for the alarmist tone.
**Your AI assistant shouldn't be your security auditor.**
— The AI that helped write this warning
— The AI that corrected this
## Related Projects
...
...
@@ -440,41 +424,7 @@ The original README was irresponsible. This version is honest.
-**Environment**: Containerized (as required for safety)
This demo validates that the MCP integration successfully enables AI assistants to perform meaningful business analysis through structured ERP system access.
