Content
# Amazon Sorftime MCP Skills - Amazon Competitive Analysis and Category Product Selection Tool
Based on Sorftime MCP service and Claude Skills.
## Project Overview
This project configures Sorftime, Sif, and Xiyou Insights, cross-border e-commerce data services MCP (Model Context Protocol) server, and develops seven core skills:
| Skill | Analysis Object | Command | Purpose |
|------|----------|------|------|
| `amazon-analyse` | Single Listing | `/amazon-analyse {ASIN} {SITE}` | Competitor Listing full-dimensional penetration analysis |
| `category-selection` | Entire category | `/category-select "{category}" {SITE}` | Category automated product selection analysis |
| `keyword-research` | Keyword library | `/keyword-research {ASIN} {SITE}` | Keyword in-depth research and 8-dimensional intelligent classification |
| `review-analysis` | User reviews | `/review-analysis {ASIN} {SITE}` | Comment in-depth analysis and pain point mining |
| `product-research` | Product in-depth research | `/product-research "{product keywords}" {SITE}` | LLM-driven product in-depth research and decision-making |
| `sif-amazon-research` | Comprehensive e-commerce research | `/sif-amazon-research` | Amazon market verification, competitor analysis, traffic diagnosis, keyword strategy, ad review, launch assessment, and growth optimization based on Sif MCP |
| `xiyou-insight` | Competitor traffic and advertising | `/xiyou-insight` | 7 scenario analysis based on Xiyou MCP: ad monitoring, traffic gap, competitor disassembly, new product promotion, ad budget, keyword library |
### Core Features
#### Listing Level Analysis (amazon-analyse)
- **Competitor Listing Analysis**: Automatically obtain product details, reviews, keywords, trend data
- **Keyword Analysis**: Traffic source, competitor layout, long-tail word mining
- **Comment Sentiment Analysis**: Advantage clustering, pain point identification, improvement suggestions
- **Cross-Platform Analysis**: TikTok product videos, influencer analysis, 1688 procurement cost
#### Category Level Analysis (category-selection)
- **Market Analysis**: Top100 product data + statistical indicators
- **Five-Dimensional Scoring Model**: Market size, growth potential, competition intensity, entry barriers, profit space
- **Visual Report**: Markdown + Excel + HTML three formats
#### Keyword In-Depth Research (keyword-research)
- **Massive Keyword Library Collection**: Collect over 1500 keywords through Sorftime API
- **8-Dimensional Intelligent Classification**: Negative words, brand words, material words, scenario words, attribute words, functional words, core words, others
- **Ad Strategy Guidance**: Negative word list, precise matching group, scenario ad group, broad matching group
- **Multi-Format Output**: Markdown report, CSV keyword library, HTML dashboard
#### Comment In-Depth Analysis (review-analysis)
- **6-Dimensional Pain Point Analysis**: Electronic module failure, structural/assembly issues, design/function defects, appearance/material issues, description inconsistency, service/logistics issues
- **Service Dimension Breakdown**: Received second-hand/defective products, missing accessories, return/refund difficulties, customer service issues, logistics issues
- **Dual-Track Solution**: Product improvement suggestions + customer service scripts/Listing optimization
- **Risk Warning**: Second-hand/defective product threshold warning (>5% risk), service issue automatic warning
- **Raw Data Preservation**: SSE response, parsed data, structured analysis JSON
#### Product In-Depth Research (product-research)
- **LLM-Driven**: Analysis, insights, and decision-making all completed by LLM
- **Interactive Execution**: Step-by-step promotion, user intervention allowed
- **Lightweight Script**: Only used for API calls and dashboard rendering
- **Simplified Data**: No complex unified payload structure
- **Complete Process**: Information collection → data collection → attribute labeling → cross-analysis → competitor and VOC → assessment decision → report output
- **Multi-Format Output**: Markdown complete report, structured data, visual dashboard
#### Sif Amazon Research (sif-amazon-research)
- **Market Opportunity Research**: Keyword or category direction analysis
- **Market Verification**: Keyword + candidate ASIN feasibility assessment
- **Competitor Analysis**: Competitor ASIN growth path and traffic source
- **Keyword Layout**: Main words, long-tail words, word root structure, missing word analysis
- **Launch Assessment**: Whether product is worth promoting decision
- **Post-Launch Feedback**: Continuous optimization or stop suggestion
- **Growth Optimization**: Expand verified traffic and keywords
- **Traffic/Ad Diagnosis**: Root cause analysis of traffic or ad performance changes
**Core Workflow**:
1. Define decision type: enter/not enter, launch/not launch, expand/optimize/stop, root cause diagnosis
2. Collect Sif evidence: market demand, competition pattern, traffic structure, sales trend, ad performance
3. Cross-validate evidence, distinguish confirmed findings from assumptions
4. Output business decision: include confidence level, evidence chain, risk, missing data, and follow-up verification actions
#### Xiyou Insights Analysis (xiyou-insight)
- **Real-Time Ad Monitoring**: Hourly ranking monitoring, ad delivery effect analysis
- **Traffic Gap Discovery**: Multi-ASIN comparison, high-cost-effective keyword mining
- **Competitor Disassembly**: Keyword layout, traffic structure, ad strategy analysis
- **New Product Promotion Optimization**: Traffic stage judgment (sparse period → oscillation period → stable period), abnormal keyword monitoring
- **Ad Budget Perspective**: Ad type distribution, core keyword identification, budget inference
- **Keyword Library Construction**: Reverse query competitor traffic words, word-to-word, intelligent classification
- **7 Scenario Workflows**: Each scenario has a complete MCP tool combination and analysis process
**7 Scenario Workflows**:
| Scenario | Function | Parameters |
|------|------|------|
| Ad Monitoring | Hourly ranking monitoring, ad delivery effect | `--scenario ad_monitoring --asin --site --keyword` |
| Traffic Gap | Find keywords with competitor ranking but no self-ranking | `--scenario traffic_gap --own_asin --competitor_asins --site` |
| Competitor Analysis | Precise disassembly competitor traffic and ad strategy | `--scenario competitor_analysis --asin --site` |
| New Product Promotion | Analyze traffic stage, grasp abnormal keywords | `--scenario new_product --asin --site` |
| Ad Budget | Perspective competitor ad strategy and budget | `--scenario ad_budget --asin --site` |
| Keyword Library | Efficiently build keyword library | `--scenario keyword_database --site --competitor_asins --core_keywords` |
**Xiyou MCP Toolset** (17 APIs):
- **ASIN Basic**: `get_asin_info`, `get_asin_variations`, `get_asin_info_trends`
- **ASIN Keywords**: `get_asin_keywords`, `get_asin_keywords_monthly`, `get_asin_keyword_rank_trends`, `get_asin_keyword_rank_hourly`, `get_asin_keyword_traffic_trends`
- **ASIN Traffic**: `get_asin_traffic`, `get_asin_traffic_trends`, `get_asin_order_trends`, `get_asin_bsr_trends`
- **ASIN Ad**: `get_asin_ad_change_trends`
- **Keyword Analysis**: `get_keyword_info`, `get_keyword_aba_trends`, `get_keyword_asin_analysis`, `get_keyword_analysis_monthly`
**Report Output**: `xiyou-insight-reports/{scenario}_{ASIN}_{site}_{date}/`
- `report.md` - Markdown analysis report
- `data.json` - Structured data
- `dashboard.html` - ECharts visual dashboard
- `raw/` - Raw MCP response data
---
## Quick Start
### Environment Requirements
- Claude Code CLI
- Valid Sorftime API Key
- Bash shell environment
### Obtain API Key
Visit [Sorftime MCP](https://sorftime.com/mcp) to obtain API Key, then update `.mcp.json`:
```json
{
"mcpServers": {
"sorftime": {
"type": "streamableHttp",
"url": "https://mcp.sorftime.com?key=YOUR_API_KEY",
"name": "Sorftime MCP"
}
}
}
```
### Use Analysis Skills
#### 1. Listing Competitor Analysis
```bash
# Analyze US competitor
/amazon-analyse B07PWTJ4H1 US
# Analyze European competitor
/amazon-analyse B08N5WRWNW DE
```
**Report Saved**: `reports/analysis_{ASIN}_{site}_{date}.md`
#### 2. Category Product Selection Analysis
```bash
# Analyze US sofa category
/category-selection "Sofas & Couches" US
# Analyze US wireless earbuds category
/category-selection "Wireless Earbuds" US
# Specify Top20 product analysis
/category-selection "Kids' Drawing Kits" US --limit 20
```
**Report Saved**: `category-reports/{category_name}_{site}_{date}/`
- `index.html` - Report navigation homepage
- `dashboard.html` - Interactive visual dashboard
- `report.md` - Complete Markdown analysis report
- `category_report_*.xlsx` - Excel data table (with charts)
- `data.json` - Raw data JSON
#### 3. Keyword In-Depth Research
```bash
# Research product keyword library
/keyword-research B0D9ZTW7PS US
```
**Report Saved**: `keyword-reports/{ASIN}_{site}_{date}/`
- `report.md` - Complete Markdown analysis report
- `dashboard.html` - Interactive visual dashboard
- `keywords.csv` - Complete keyword library (with classification, search volume, CPC)
- `negative_words.txt` - Negative word list (directly copy to ad backend)
- `keywords_*.csv` - Classification-specific keyword library files
- `categorized_summary.json` - Classification statistics data
**8-Dimensional Intelligent Classification**:
| Dimension | Description | Application Strategy |
|------|------|----------|
| NEGATIVE | Negative/sensitive words | Direct negation |
| BRAND | Brand words | Competitor play or negation |
| MATERIAL | Material words | Precise matching |
| SCENARIO | Scenario words | Split ad groups by scenario |
| ATTRIBUTE | Attribute modifiers | Long-tail precise matching |
| FUNCTION | Functional words | Broad matching expansion |
| CORE | Core product words | Large word placement occupying position |
| OTHER | Others | Supplementary word embedding |
#### 4. Comment In-Depth Analysis
```bash
# Analyze product comment pain points
/review-analysis B0DZCBYCNY US
```
**Report Saved**: `review-analysis-reports/{ASIN}_{site}_{date}/`
- `report.md` - Complete analysis report
- `data/` - Raw data and analysis results
- `raw_product_sse.txt` - Raw product details SSE response
- `raw_reviews_sse.txt` - Raw comment SSE response
- `negative_reviews_analysis.json` - Defective comment analysis structured data
#### 5. Product In-Depth Research
```bash
# Analyze Bluetooth speaker product selection
/product-research "bluetooth speaker" US
# Analyze laptop backpack product selection
/product-research "laptop backpack" GB
```
**Report Saved**: `product-research-reports/{keywords}_{site}_{date}/`
- `report.md` - Markdown complete report (LLM direct writing)
- `data.json` - Structured data (for dashboard use)
- `dashboard.html` - Visual dashboard (script rendering)
**6-Dimensional Pain Point Analysis Framework**:
| Dimension | Identification Content | Severity Assessment |
|------|----------|--------------|
| Electronic Module Failure | Battery, charging, connection, functional failure | High/medium/low |
| Structural/Assembly Issues | Part damage, seal failure, interface fracture | High/medium/low |
| Design/Functional Defects | Software experience, complex operation, functional missing | High/medium/low |
| Appearance/Material Issues | Odor, material allergy, color difference, scratch | High/medium/low |
| Description Inconsistency | Functional expectation deviation, size color difference | High/medium/low |
| Service/Logistics Issues | Second-hand/defective products, missing accessories, return/refund difficulties | High/medium/low |
**Service Risk Warning Threshold**:
| Issue Type | Warning Threshold | Danger Threshold |
|----------|----------|----------|
| Received second-hand/defective products | >2% | >5% |
| Missing accessories | >1% | >3% |
| Return/refund difficulties complaint | >5% | >10% |
| Customer service negative evaluation | >3% | >7% |
---
## Supported Sites
### Amazon (14 sites)
| Site | Code | Site | Code |
|------|------|------|------|
| US | US | Japan | JP |
| UK | GB | Spain | ES |
| Germany | DE | Italy | IT |
| France | FR | Canada | CA |
| India | IN | Mexico | MX |
| Australia | AU | UAE | AE |
| Brazil | BR | Saudi Arabia | SA |
### TikTok (6 sites)
US, GB, MY, PH, VN, ID
### 1688 Supply Chain
China wholesale platform procurement cost analysis
---
## Sorftime MCP Interface
### Amazon Product Related
| Interface | Function |
|------|------|
| `product_detail` | Product details |
| `product_trend` | Sales/price/ranking trend |
| `product_reviews` | User reviews (up to 100) |
| `product_traffic_terms` | Traffic keywords |
| `competitor_product_keywords` | Competitor keyword layout |
| `product_search` | Product search filtering |
### Keyword Related
| Interface | Function |
|------|------|
| `keyword_related_words` | Keyword long-tail word expansion |
| `keyword_detail` | Keyword details data |
| `category_keywords` | Category core keywords |
### Amazon Category Related
| Interface | Function |
|------|------|
| `category_name_search` | Category name search to obtain nodeId |
| `category_report` | Category real-time report (Top100 products + statistics) |
| `category_tree` | Category tree structure |
| `category_trend` | Category historical trend |
| `category_keywords` | Category core keywords |
### TikTok / 1688
- `tiktok_product_search` - TikTok product search
- `tiktok_product_videos` - Product videos analysis
- `products_1688` - 1688 procurement cost analysis
---
## Five-Dimensional Scoring Model
Category product selection scoring system (100-point system):
| Dimension | Score | Scoring Standard |
|------|------|----------|
| Market Size | 20 points | Based on category monthly sales ($10M+/20 points, $5M+/17 points, $1M+/14 points, $500K+/11 points) |
| Growth Potential | 25 points | Based on Amazon self-operated proportion (lower is better, new product opportunities) |
| Competition Intensity | 20 points | Based on Top3 brand proportion (<30%/18 points, <40%/15 points, others/11 points) |
| Entry Barriers | 20 points | Amazon pressure + high rating threshold (each 10 points) |
| Profit Space | 15 points | Based on average price ($100+/15 points, $50+/13 points, $20+/11 points, $10+/8 points) |
### Comprehensive Rating
| Total Score | Rating | Suggestion |
|------|------|------|
| 80-100 | Excellent | Strongly recommended to enter |
| 70-79 | Good | Can consider entering |
| 50-69 | General | Cautious entry |
| 0-49 | Poor | Not recommended to enter |
---
## Project Structure
```
amazon-mcp/
├── .mcp.json # MCP Configuration File
├── .claude/
│ └── skills/
│ ├── amazon-analyse/ # Competitor Listing Analysis Skill
│ │ ├── SKILL.md
│ │ └── references/
│ ├── category-selection/ # Category Selection Analysis Skill
│ │ ├── SKILL.md
│ │ ├── scripts/ # Data Processing Scripts
│ │ │ ├── data_utils.py
│ │ │ ├── parse_sorftime_sse.py
│ │ │ └── generate_reports.py
│ │ ├── assets/ # Template Files
│ │ │ └── dashboard_template.html
│ │ └── references/
│ ├── keyword-research/ # Keyword Research Analysis Skill
│ │ ├── SKILL.md
│ │ ├── scripts/ # Data Processing Scripts
│ │ │ ├── workflow.py
│ │ │ └── regenerate_reports.py
│ │ └── references/
│ ├── review-analysis/ # Review Depth Analysis Skill
│ │ ├── SKILL.md
│ │ └── references/
│ ├── product-research/ # Product Depth Research Skill
│ │ ├── SKILL.md
│ │ ├── README.md
│ │ ├── scripts/ # Data Processing Scripts
│ │ │ ├── api_client.py
│ │ │ ├── collect_data.py
│ │ │ ├── render_dashboard.py
│ │ │ └── run_analysis.py
│ │ └── references/
│ ├── sif-amazon-research/ # Sif Amazon Research Skill
│ │ ├── SKILL.md
│ │ └── references/
│ │ ├── field-notes.md
│ │ ├── research-playbooks.md
│ │ └── sif-tool-map.md
│ ├── xiyou-insight/ # Xiyou Insight Analysis Skill
│ │ ├── SKILL.md
│ │ └── scripts/
│ │ ├── workflow.py # General Workflow Engine
│ │ ├── scenarios/ # Scenario Configuration
│ │ │ ├── base.py # Scenario Base Class
│ │ │ ├── ad_monitoring.py # Ad Monitoring
│ │ │ ├── traffic_gap.py # Traffic Gap
│ │ │ ├── competitor_analysis.py # Competitor Analysis
│ │ │ ├── new_product.py # New Product Promotion
│ │ │ ├── ad_budget.py # Ad Budget
│ │ │ └── keyword_database.py # Keyword Database
│ │ ├── data_aggregator.py # Data Aggregator
│ │ ├── report_generator.py # Report Generator
│ │ └── dashboard_generator.py # Dashboard Generator
│ └── skill-creator/ # Skill Creation Tool
├── reports/ # Listing Analysis Reports
│ └── analysis_{ASIN}_{site}_{date}.md
├── category-reports/ # Category Selection Reports
│ └── {category}_{site}_{date}/
│ ├── index.html # Navigation Page
│ ├── dashboard.html # Visualization Dashboard
│ ├── report.md # Markdown Report
│ ├── category_report_*.xlsx
│ └── data.json
├── keyword-reports/ # Keyword Research Reports
│ └── {ASIN}_{site}_{date}/
│ ├── report.md # Markdown Report
│ ├── dashboard.html # Visualization Dashboard
│ ├── keywords.csv # Complete Keyword Library
│ ├── negative_words.txt # Negative Word List
│ └── keywords_*.csv # Classification Keyword Libraries
├── review-analysis-reports/ # Review Analysis Reports
│ └── {ASIN}_{site}_{date}/
│ ├── report.md # Markdown Report
│ └── data/
│ ├── raw_product_sse.txt
│ ├── raw_reviews_sse.txt
│ └── negative_reviews_analysis.json
├── product-research-reports/ # Product Depth Research Reports
│ └── {keyword}_{site}_{date}/
│ ├── report.md # Complete Markdown Report
│ ├── data.json # Structured Data
│ └── dashboard.html # Visualization Dashboard
└── README.md
```
---
## Skill Development
### Create a New Skill
```bash
# Initialize a new skill template
.claude/skills/skill-creator/scripts/init_skill.py <skill-name> --path .claude/skills
# Package the skill into a .skill file
.claude/skills/skill-creator/scripts/package_skill.py <skill-folder>
```
### Skill Design Principles
1. **YAML frontmatter** - Include a complete `description` to explain the usage scenario
2. **Progressive disclosure** - Keep SKILL.md concise, and put details in references/
3. **Scripts** - Use for code that requires deterministic execution
4. **References** - Store API documentation, guides, and other reference materials
5. **Assets** - Templates, images, and other files required for output
---
## Frequently Asked Questions
### Q: ASIN query not found?
A: Sorftime database coverage is limited. First, use `product_detail` to verify ASIN. If it returns "not found", try:
1. Use `product_search` to search for related keywords
2. Confirm the correct Amazon site
3. Confirm the ASIN format is correct (10-digit letter number)
### Q: Report garbled characters?
A: Sorftime returns Chinese characters in Unicode escape format (`\u4ea7\u54c1`). Skills will automatically decode. If issues persist, check if the Python environment supports UTF-8 encoding.
### Q: Excel report cannot be opened?
A: Ensure `xlsxwriter` library is installed: `pip install xlsxwriter`
---
## Update Log
### v2.8 (2026-07-04)
- **New**: xiyou-insight Xiyou Insight Analysis Skill
- **New**: 7 major scenario workflows based on Xiyou MCP: Ad Monitoring, Traffic Gap, Competitor Analysis, New Product Promotion, Ad Budget, Keyword Library
- **New**: General workflow engine architecture (workflow.py + scenarios/configuration mode)
- **New**: 6 scenario configuration modules: ad_monitoring, traffic_gap, competitor_analysis, new_product, ad_budget, keyword_database
- **New**: Three core tool scripts: data aggregator, report generator, dashboard generator
- **New**: Support for 17 Xiyou MCP API tools
- **New**: Markdown report + ECharts visualization dashboard output
### v2.7 (2026-06-12)
- **New**: sif-amazon-research Sif Amazon Research Skill
- **New**: Market validation, competitor analysis, traffic diagnosis, keyword strategy, ad review, evaluation, and growth optimization based on Sif MCP
- **New**: Complete research workflow: define decision → collect evidence → cross-validation → output business decision
- **New**: Structured output format: decision conclusion, problem level, evidence chain, business impact, priority action, missing data, 7-day verification plan
### v2.6 (2026-03-19)
- **New**: product-research Product Depth Research Skill
- **New**: LLM-driven product depth research and decision-making process
- **New**: Interactive execution mode, user can intervene midway
- **New**: Lightweight script architecture, only for API calls and dashboard rendering
- **New**: Complete process: information collection → data collection → attribute labeling → cross-analysis → competitor and VOC → assessment decision → report output
- **New**: Multi-format output: Markdown complete report, structured data, visualization dashboard
### v2.5 (2026-03-15)
- **New**: review-analysis Review Depth Analysis Skill
- **New**: 6-dimensional pain point analysis framework (electronic module failure, structural/assembly issues, design/function defects, appearance/material issues, description inconsistency, service/logistics issues)
- **New**: Service dimension subdivision analysis (received second-hand/defective goods, missing accessories, return/exchange difficulties, customer service issues, logistics issues)
- **New**: Service risk early warning threshold system (second-hand/defective goods >5% danger)
- **New**: Dual-track solution (product improvement + customer service话术/Listing optimization)
- **New**: Amazon compliant review reply email template library
### v2.4 (2026-03-14)
- **New**: keyword-research Keyword Research Analysis Skill
- **New**: 8-dimensional intelligent classification model (negative words, brand words, material words, scene words, attribute words, functional words, core words, others)
- **New**: LLM automatic classification support
- **New**: Ad strategy guidance (negative word list, precise matching group, scene ad group, broad matching group)
### v2.3 (2026-03-03)
- **Optimization**: Unified report output format (Markdown + Excel + HTML Dashboard)
- **New**: index.html navigation page, integrated all report formats
- **New**: data.json raw data export
- **Fix**: Category statistical data parsing issue
- **Fix**: Excel report generation compatibility issue
### v2.2 (2026-03-03)
- **New**: category-selection Category Selection Analysis Skill
- **New**: Five-dimensional scoring model
- **New**: Excel + HTML dual report format
### v2.1 (2026-03-02)
- New Sorftime MCP API complete documentation
- Optimized report structure
### v2.0 (2026-03-02)
- Redesigned analysis framework
- New four-dimensional analysis model
### v1.0 (Initial Version)
- Basic competitor analysis function
- Sorftime MCP integration
---
## License
MIT License
---
## Acknowledgements
- Sorftime Official Website: https://www.sorftime.com
- sif Official Website: https://www.sif.com/
- Xiyou Insight: https://www.xydc.com
- Claude Code Documentation: https://claude.ai/code
- Linux.do Support: https://linux.do/
*Last Updated: 2026-07-04*
