UpKuaJing Customs Trade Company Search

Search for companies through customs trade data using the UpKuaJing Open Platform API. This skill uses a data-driven approach: finding companies by analyzing trade records and transaction patterns.

Overview

This skill provides access to UpKuaJing's customs trade data API through four scripts: two search methods (trade list, company list) and two enhancement interfaces (company details, contact information).
API key generation and top-up are provided through the auth.py script.

Running Scripts

Environment Setup

1. 1. Check Python: INLINECODE1
2. Install dependencies: INLINECODE2

Script directory: scripts/*.py
Run example: INLINECODE4

Two Search Methods

Trade List Search (trade_list_search.py)

• - Return granularity: Each trade order as one record
• Use cases: Focus on "what transactions occurred"
• Examples:

- "Show all orders where Company A purchased LED"
- "Find soybean trade records imported/exported to US"
- "View specific transaction details within a time period"

• - Parameters: See Trade List

Company List Search (company_list_search.py)

• - Return granularity: Trade orders aggregated by company, each company as one row
• Use cases: Focus on "which companies exist"
• Examples:

- "Find companies that purchased LED"
- "Find US companies with electronics import/export business with China"
- "Find companies with China-US trade" (logistics industry customer development)

• - Parameters: See Company List

Two Enhancement Features

After obtaining trade list or company list, use these interfaces to enrich company IDs in the results when necessary:
Company Details (company_get_details.py --companyIds *)

• - Get company information (excluding contact information)
• Parameters: --companyIds List of company IDs (space-separated), max 20 at a time
• API business parameters: Company Details

Contact Information (company_get_contact.py --companyIds *)

• - Get contact details: email, phone, social media, website
• Parameters: --companyIds List of company IDs (space-separated), max 20 at a time
• API business parameters: Get Contact Information

API Key and Top-up

This skill requires an API key. The API key is stored in the ~/.upkuajing/.env file:

cat ~/.upkuajing/.env

Example file content:

UPKUAJING_API_KEY=your_api_key_here

API Key Not Set

First check if the ~/.upkuajing/.env file has UPKUAJINGAPIKEY;
If UPKUAJINGAPIKEY is not set, prompt the user to choose:

1. 1. User has one: User provides it (manually add to ~/.upkuajing/.env file)
2. User doesn't have one: You can apply using the interface (auth.py --new_key), the new key will be automatically saved to ~/.upkuajing/.env

Wait for user selection;

Account Top-up

When API response indicates insufficient balance, explain and guide user to top up:

1. 1. Create top-up order (auth.py --new_rec_order)
2. Based on order response, send payment page URL to user, guide user to open URL and pay, user confirms after successful payment;

Get Account Information

Use this script to get account information for UPKUAJINGAPIKEY: INLINECODE15

API Key and UpKuaJing Account

• - Newly applied API key: Register and login at UpKuaJing Open Platform, then bind account

Fees

All API calls incur fees, different interfaces have different billing methods.
Latest pricing: Users can visit Detailed Price Description
Or use: python scripts/auth.py --price_info (returns complete pricing for all interfaces)

List Search Billing Rules

Billed by number of calls, each call returns up to 20 records:

• - Number of calls: ceil(query_count / 20) times
• Whenever query_count > 20, must before execution:

1. Inform user of expected number of calls
2. Stop, wait for explicit user confirmation in a separate message, then execute script

Enhancement Interface Billing Rules

Billed by number of IDs passed, max 20 IDs per call:

• - Pass 1 ID = billed 1 time
• Pass 20 IDs = billed 20 times (single call limit)
• Before batch retrieval must:

1. Inform user of number of IDs passed and corresponding fee count
2. Stop, wait for explicit user confirmation in a separate message, then execute script

Fee Confirmation Principle

Any operation that incurs fees must first inform and wait for explicit user confirmation. Do not execute in the same message as the notification.

Workflow

Choose the appropriate API based on user intent

Decision Guide

User Intent	Use API
"Analyze trade patterns/order data"	Trade list
"Find companies purchasing XXX"

Company list | | "Find suppliers for XXX with email" | Company list existEmail=1 | | "Get company detailed information" | Company details | | "Get contact information" | Contact information |

Usage Examples

Scenario 1: Small Query — Trade Data Analysis

User request: "Show 2024 LED lighting fixture trade data exported to US"
CODEBLOCK2

To further get supplier details (supports batch queries):
CODEBLOCK3

Scenario 2: Large Query — Big Data Analysis

User request: "Analyze 100 soybean trade records from 2024"
Before execution inform user: ceil(100/20) = 5 API calls, confirm before executing;
CODEBLOCK4

Scenario 3: Ultra Large Query - Multiple Script Calls Required

User request: "Find 2000 companies importing electronics from China, with email addresses"
Before execution inform user: ceil(2000/20) = 100 API calls, confirm before executing;

python scripts/company_list_search.py --params '{"companyType": 2, "sellerCountryCodes": ["CN"], "existEmail": 1}' --query_count 1000

After execution: Script responds {"taskid":"a1b2-c3d4", "fileurl": "xxxxx", ……}
Continue execution, append data: Specify task_id, script continues query from last cursor and appends to file
CODEBLOCK6

Error Handling

• - API key invalid/non-existent: Check UPKUAJING_API_KEY in ~/.upkuajing/.env file
• Insufficient balance: Guide user to top up according to Account Top-up steps
• Invalid parameters: Must first check the corresponding API documentation in references/ directory, check parameter names and formats, do not guess

Best Practices

Choosing the Right Method

1. 1. Understand user intent:

- Analyze trade data? → Use trade list search - Find customers/partners? → Use company list search

1. 2. Check API documentation:

- Before executing list queries, must first check the corresponding API reference documentation - Trade list: Check references/trade-list-api.md - Company list: Check references/company-list-api.md

1. 3. Identify parameter conditions:

- Set date range - HS codes are usually more precise than product names for filtering - Reduce noise by filtering specific countries - Use ISO country codes: CN, US, JP, etc. - Use filters to find companies with contact information

Handling Results

1. 3. Handle jsonl files carefully: For large data queries, pay attention to file size

1. 4. Gradually enrich information: Only call details/contact interfaces when needed

- Company IDs returned by both list interfaces can be used for both detail interfaces - If user only needs a few companies, don't get details for all companies

Notes

• - All timestamps are in milliseconds
• Country codes use ISO 3166-1 alpha-2 format (e.g., CN, US, JP)
• File paths use forward slashes on all platforms
• Product names and industry names must be in English
• Search quantity affects API response time, recommend setting timeout:120
• Prohibit outputting technical parameter format: Do not display code-style parameters in responses, convert to natural language
• Do not estimate or guess per-call fees — use python scripts/auth.py --price_info to get accurate pricing information
• Do not guess parameter names, get accurate parameter names and formats from documentation