chip-trade-skills

IC Trade Navigator — MCP Connector

Taiwan-neutral IC component quoting connector for Claude Desktop and MCP-compatible clients.

The bridge between your local ERP inventory and real-time IC market intelligence.

Connect your local ERP inventory to real-time market intelligence — without sending your pricing data to any server.

What This Is

This connector is the client-side component of the IC Trade Navigator system. It:

Your floor prices, purchase costs, and ERP financial data never leave your machine.

🛡️ Why Use This? Privacy First: Your cost and financial data never leave your local machine.

Market Intelligence: Real-time pricing and risk scoring (low/medium/high).

Taiwan-Neutral: Specialized filtering and advisory for the Taiwan electronics trade.

Architecture

Your Machine                          IC Trade Navigator Server
─────────────────────────────         ──────────────────────────
inventory.xlsx  ←─ read               (Black-box API)
    │                                  • Market scraping
    │  part_number + qty only          • Risk scoring
    └──────────────────────────────►   • Multilingual advisory
                                       • Taiwan-neutral filtering
    ◄──────────────────────────────
    quoted_price, risk_level,
    advisory (en/de/ja/zh-TW/fr/ko)
         │
    merged_view  ─► Claude Desktop

Quick Start

1. Install dependencies

pip install httpx openpyxl

2. Configure environment

cp .env.example .env
Edit .env:
  QUOTE_ENGINE_URL     = https://api.ic-navigator.com   (or your self-hosted URL)
  QUOTE_ENGINE_API_KEY = JAY-IC-xxxxxxxxxxxxxxxxxxxx
  ERP_EXCEL_PATH       = /path/to/your/inventory.xlsx

3. Single quote (CLI)

python -m mcp_connector.client quote STM32L412CBU6 --qty 500 --lang zh-TW

4. Batch quote from inventory

python -m mcp_connector.client batch data/inventory.xlsx

5. JSON output (for ERP integration)

python -m mcp_connector.client quote GD32F103C8T6 --qty 1000 --lang en --json

MCP Tools (Claude Desktop)

Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "ic-trade-navigator": {
      "command": "python",
      "args": ["-m", "mcp_connector.server"],
      "env": {
        "QUOTE_ENGINE_URL": "https://api.ic-navigator.com",
        "QUOTE_ENGINE_API_KEY": "JAY-IC-your-key-here",
        "ERP_EXCEL_PATH": "/absolute/path/to/inventory.xlsx"
      }
    }
  }
}

Available tools in Claude:

| Tool | Description | |---|---| | quote_part | Get market quote + risk score for one part number | | read_erp_inventory | Look up your local stock for a part | | get_combined_view | Full merged view: market + local ERP |

Inventory File Format

Your inventory.xlsx should have these columns (column names are configurable):

| Part Number | Stock Qty | Status | Package | Date Code | |---|---|---|---|---| | STM32L412CBU6 | 12000 | In Stock | UFQFPN32 | 2347 | | GD32F103C8T6 | 8000 | In Stock | LQFP48 | 2344 |

A sample file is included at data/inventory.xlsx.

Pricing columns are automatically blocked — even if present in your file, the connector will never read or transmit them.

Response Fields

{
  "part_number": "STM32L412CBU6",
  "quoted_price": 2.8500,
  "quote_action": "auto_quote",
  "risk_level": "low",
  "risk_index": 0.12,
  "tw_neutral_confidence": 0.88,
  "advisory": "Part is in normal active supply...",
  "advisory_lang": "en",
  "local_stock_qty": 12000,
  "recommendation": "✅ Auto-quote ready. Local stock: 12,000 units."
}

Risk Levels

| Level | Score | Meaning | |---|---|---| | 🟢 low | < 0.30 | Standard procurement confidence | | 🟡 medium | 0.30–0.65 | Request Certificate of Conformance | | 🔴 high | ≥ 0.65 | Escalate to procurement quality team |

Advisory Languages

en · de · ja · zh-TW · fr · ko

Custom Column Mapping

If your Excel uses different column headers:

python -m mcp_connector.client batch inventory.xlsx \
  --col-pn "MPN" \
  --col-qty "Available" \
  --col-status "Lifecycle"

Environment Variables

| Variable | Required | Default | Description | |---|---|---|---| | QUOTE_ENGINE_URL | ✅ | — | API server base URL | | QUOTE_ENGINE_API_KEY | ✅ | — | Your JAY-IC- API key | | ERP_EXCEL_PATH | — | data/inventory.xlsx | Path to local inventory file | | CONNECTOR_TIMEOUT | — | 15 | HTTP timeout in seconds |

Privacy Guarantee

This connector enforces a strict data boundary:

License

MIT License — see LICENSE

Get an API Key

This connector requires access to the IC Trade Navigator API. Contact: jay.ictw@gmail.com

1. Install dependencies

pip install httpx openpyxl

2. Configure environment

cp .env.example .env
Edit .env:
  QUOTE_ENGINE_URL     = https://api.ic-navigator.com   (or your self-hosted URL)
  QUOTE_ENGINE_API_KEY = JAY-IC-xxxxxxxxxxxxxxxxxxxx
  ERP_EXCEL_PATH       = /path/to/your/inventory.xlsx

3. Single quote (CLI)

python -m mcp_connector.client quote STM32L412CBU6 --qty 500 --lang zh-TW

4. Batch quote from inventory

python -m mcp_connector.client batch data/inventory.xlsx

5. JSON output (for ERP integration)

python -m mcp_connector.client quote GD32F103C8T6 --qty 1000 --lang en --json