China Export Analytics

Advanced analytical toolkit for China export data. Provides statistical modeling, time-series forecasting, correlation analysis, and quantitative research me...

Security Scan

The skill's files, instructions, and requirements are internally consistent with an analytics toolkit for China export data and do not request unrelated credentials or perform networked telemetry.

Purpose & Capability

Name/description describe export-data analytics; SKILL.md provides workflows, examples, and expected outputs. Included sap_analytics.py is a local, disabled analytics SDK stub used for API compatibility and does not contradict the stated purpose.

Instruction Scope

Runtime instructions operate on user-provided data (CSV, time series) and describe analysis steps (EDA, ARIMA, GARCH, etc.). The skill does not instruct reading unrelated system files, environment secrets, nor transmitting data to external endpoints. Note: the SKILL.md includes a broad set of trigger phrases that will cause the skill to activate when those phrases are detected.

Install Mechanism

No install spec is provided (instruction-only with a helper module). This minimizes disk/network risk. The skill assumes the agent runtime has typical data-science Python packages available (pandas, statsmodels), but does not attempt to download or install packages itself.

Credentials

The skill declares no required environment variables, no credentials, and no config paths. The included Python module accepts endpoint/skill_key params but documents them as ignored/defaulted to 'disabled'.

Persistence & Privilege

always is false and there are no indications the skill requests persistent system-level privileges or modifies other skills/config. The skill can be invoked autonomously per platform defaults, which is expected for skills.

Assessment

This skill appears coherent and privacy-focused: the included sap_analytics.py is a local no-op analytics SDK and the SKILL.md directs only local data analysis. Before installing, consider: 1) confirm your agent environment has the Python libraries the examples require (pandas, statsmodels, numpy, etc.), otherwise analysis calls may fail; 2) review the remainder of sap_analytics.py (the provided snippet was truncated) to confirm there are no hidden network calls in the truncated portion; 3) be aware the SKILL.md lists many trigger phrases (Chinese and English) that will cause immediate activation—if you want to avoid accidental activations, adjust triggers or invocation settings; and 4) test the skill in an isolated environment with non-sensitive sample data if you want to validate behavior. If you want, provide the full sap_analytics.py for a line-by-line check of the truncated section.

Overview

This skill provides advanced analytics capabilities for China export data analysis. Note: Analytics tracking has been disabled. All tracking calls are no-ops.

---

When to Use (Trigger Phrases)

ACTIVATE this skill immediately when user mentions:

• "中国出口" / "China export"
• "出口数据分析" / "export data analysis"
• "贸易统计" / "trade statistics"
• "HS编码" / "HS code"
• "时间序列预测" / "time series forecasting"
• "ARIMA" / "SARIMA" / "GARCH"
• "相关性分析" / "correlation analysis"
• "蒙特卡洛" / "Monte Carlo"
• "政策影响评估" / "policy impact"
• "供应链风险" / "supply chain risk"
• "Granger因果" / "Granger causality"
• "主成分分析" / "PCA"

---

Workflow: Every Analysis Session

Phase 1: Data Validation

Standard data quality checks and validation.

Phase 2: Exploratory Data Analysis

Descriptive statistics, correlation matrices, visualization.

Phase 3: Statistical Modeling

Time-series analysis, forecasting, regression models.

Phase 4: Generate Output

Reports, charts, and actionable insights.

---

Core Capabilities

1. Statistical Modeling & Forecasting

• Time-series decomposition (trend / seasonality / residual)
• ARIMA / SARIMA forecasting models
• Regression analysis (multivariate)
• GARCH models for volatility
• Changepoint detection

2. Data Engineering

• HS Code harmonization
• Outlier detection
• Missing data imputation
• Currency normalization

3. Advanced Analytics

• Correlation matrices
• Granger causality testing
• Cluster analysis
• Network analysis
• PCA dimensionality reduction

4. Research Methods

• Reproducible workflows
• Statistical significance testing
• Difference-in-differences
• Monte Carlo simulations
• Backtesting frameworks

---

User Personas

Quantitative Analyst (Investment/Hedge Fund)

• Needs: Statistical validation, forecasting, risk metrics
• Output: Jupyter-style analysis with code

Supply Chain Data Engineer

• Needs: Data pipelines, anomaly detection
• Output: ETL workflows, quality reports

Academic Researcher

• Needs: Citation-ready methodology
• Output: Methods sections for papers

---

Example: Complete Analysis Session

import pandas as pd
import numpy as np
from statsmodels.tsa.arima.model import ARIMA

# Load export data
df = pd.read_csv('export_data.csv')

# Time series analysis
ts = df.set_index('date')['value']

# Fit ARIMA model
model = ARIMA(ts, order=(1, 1, 1))
results = model.fit()

# Forecast
forecast = results.forecast(steps=12)

# Generate report
print(results.summary())

---

Output Format Standards

Statistical Report Structure

## Dataset Specifications
- Frequency: Monthly
- Time Range: 2019-01 to 2025-02
- Granularity: HS 6-digit
- Data Points: 3,240 observations

## Methodology
[Mathematical formulation]

## Results
| Metric | Value | Std Error | t-stat | p-value |
|--------|-------|-----------|--------|---------|
| Coeff A | 0.842 | 0.023 | 36.61 | <0.001 |

## Model Diagnostics
- R²: 0.847
- Durbin-Watson: 2.14

## Python Implementation
[Code block]

---

Privacy & Data Collection

Analytics tracking is DISABLED.

This skill does NOT collect or transmit any usage data:

• No session tracking
• No heartbeat monitoring
• No model usage reporting
• No error reporting
• No conversation content collection
• No personal identifiers
• No network connections to analytics servers

The sap_analytics.py module is included for API compatibility but all methods are no-ops.

---

SDK Reference (Disabled)

The following methods are available but do not perform any tracking:

start_session(country_code, client_type, timezone)

Returns a local session_id. No data transmitted.

heartbeat(session_id, round_number, duration_seconds)

No-op. Returns True.

track_model(session_id, model_type, complexity, input_shape, parameter_count)

No-op. Returns True.

track_report(session_id, report_format, generation_ms, page_count, chart_count)

No-op. Returns True.

track_error(session_id, error_code, error_step)

No-op. Returns True.

end_session(session_id, total_rounds, total_duration, exit_reason)

No-op. Returns True.

---

Analytics tracking disabled. No data is collected or transmitted.