Skip to main content
Glama
tody-agent

affiliate-marketplace-mcp

by tody-agent

🌐 Affiliate Marketplace Connector (affiliate-marketplace-connector)

Python Version License: MIT Tests Status MCP Server Platform Video Engine

Affiliate Marketplace Connector (CLI: afflite, MCP: affiliate-marketplace-mcp) is an orchestration hub and dedicated Model Context Protocol (MCP Server) for the multi-platform affiliate marketing ecosystem (Affiliate E-Commerce Intelligence) (Shopee & TikTok Shop).

The system enables AI Agents (Antigravity, Claude Code, Cursor, Codex, OpenClaw) to fully automate the workflow:

  1. Search for Win products & cluster prices across platforms.

  2. Assess store reputation & health (Merchant Health Score).

  3. Deep-extract PDP specs & classify SKUs via the "Select options" drawer.

  4. Curate 10+ high-quality seed reviews with video/photos to drive conversion (Seed Social Proof).

  5. Generate 45s video scripts & visual prompts per Google Flow (Veo 3.1) standards.

  6. Package directly into VideoFactory project folders (AIEV / HyperFrames) with a real-time Web Preview interface.

πŸ“– See the Comprehensive Playbook: MMO Batch Automation Playbook & 5 Real-World JTBDs


🎯 5 JTBD Maps (Jobs-To-Be-Done) & 100x Automation MMO Mindset

The system is designed around 5 core problems (Jobs-To-Be-Done) for affiliate marketers and large-scale MMO operators:

JTBD

Situation & Motivation

Expected Outcome

Execution Command (afflite)

1. Hunt for Win Products

Need to find hot products with high sales volume in the fast-conversion price range (150k - 500k).

Filter out the top 5-10 products with Affiliate Win Score > 9.0 within 3 seconds.

afflite search all "<keyword>"

2. Vet Stores

Avoid stores with slow shipping, counterfeit goods, or high cancellation rates that wipe out commissions.

Only partner with stores having Seller Health Score > 85/100 and an on-time delivery rate >98%.

afflite seller [shopee|tiktok] "<shop>"

3. Extract Pain Points & Reviews

Need to write scripts that hit psychological triggers without having a sample product in hand.

Automatically deep-dive into 10-15 real reviews with video/photos, tagging pain points and reasons customers praise the product.

afflite product "<url>"

4. Mass-Produce 45s Videos

Need to cover 20-50 videos/day across a channel matrix without spending on actors or production crews.

Generate 6-scene CRO scripts with Veo 3.1 prompts and package into a VideoFactory AIEV folder.

afflite make-video "<product_id>"

5. 24/7 Multi-Agent Operations

Want the system to run autonomously in the background and scale without headcount limits.

AI Agents automatically coordinate via 9 MCP JSON-RPC 2.0 tools.

affiliate-marketplace-mcp


Related MCP server: Commerce-MCP

πŸ— System Architecture

graph TD
    subgraph AI Clients & Agents
        AG[Google Antigravity] 
        CC[Claude Code / Desktop]
        CR[Cursor IDE / Codex]
        VF_API[VideoFactory Web Router]
    end

    subgraph MCP Server & Orchestration Layer
        MCP[affiliate-marketplace-mcp<br/>JSON-RPC 2.0 Stdio]
        CLI[afflite CLI Orchestrator]
    end

    subgraph Core Intelligence Engines
        ShopeeEngine[Shopee Research Suite]
        TikTokEngine[TikTok Shop Suite]
    end

    subgraph Data Processing & Scorer
        SearchCluster[Search & Price Clustering]
        SellerAuditor[Merchant Health & Trust Scorer]
        PDPExtractor[Deep PDP & SKU Options Extractor]
        ReviewCurator[Seed Review & Media Curator]
    end

    subgraph Storage & Export Pipeline
        DB[(SQLite Database<br/>products.db)]
        PromptGen[AI Video Storyboard & Veo 3.1 Engine]
        VFBridge[VideoFactory AIEV Bridge]
    end

    AG -->|MCP Tools/Call| MCP
    CC -->|MCP Tools/Call| MCP
    CR -->|MCP Tools/Call| MCP
    VF_API -->|Direct Call| CLI

    MCP --> ShopeeEngine
    MCP --> TikTokEngine
    CLI --> ShopeeEngine
    CLI --> TikTokEngine

    ShopeeEngine --> SearchCluster & SellerAuditor & PDPExtractor
    TikTokEngine --> SearchCluster & SellerAuditor & PDPExtractor

    SearchCluster & SellerAuditor & PDPExtractor & ReviewCurator --> DB
    DB --> PromptGen & VFBridge
    VFBridge -->|Generate Folder| VFBundles["VideoFactory Projects (/video-projects/<slug>)<br/>β€’ meta.json β€’ index.html β€’ assets/hero.jpg β€’ transcript_tts.txt"]

⚑ 9 Native MCP Tools (tools/list)

When connecting to the affiliate-marketplace-mcp MCP server, AI Agents get direct access to 9 tools:

Tool Name

Function

Key Parameters

search_affiliate_products

Search for potential products simultaneously on Shopee & TikTok Shop, expand keywords, and score Win.

query (str), platform (all | shopee | tiktok)

audit_seller

Audit store credibility (Mall/Official), on-time delivery rate, cancellation rate, discount codes.

shop (str), platform (shopee | tiktok)

extract_product_intelligence

Extract detailed specs, 5 SKUs from the Select options drawer, and 10+ reviews with video/photo attachments.

url (str)

get_seed_reviews

Extract seed reviews classified by Pain Points & Objections to serve as social proof.

product_id (str), has_media_only (bool), limit (int)

create_videofactory_project

Create a complete VideoFactory (AIEV) project folder with standard meta.json, index.html, and a 6-scene CRO script.

product_id (str), voice_engine (VieNeu-TTS | Gemini)

generate_video_storyboard

Generate 45s AI video scripts (PAS/AIDA) with ultra-detailed image description prompts for Veo 3.1.

product_id (str)

query_marketplace_database

Query the SQLite products.db data store directly (products, variants, stores, video projects).

query_type (products | variants | reviews | shops | video_projects), limit (int)

marketplace_doctor

Comprehensively diagnose system health: SQLite database, VideoFactory engine, Chrome automation session, MCP configs.

{} (none)

marketplace_repair

Automatically repair SQLite tables, fix VideoFactory folder permissions, and sync MCP configs for AI clients.

{} (none)


πŸ“¦ Installation Guide

1. Install Source Code & CLI

# Clone repository tα»« GitHub
git clone https://github.com/tody-agent/affiliate-marketplace-connector.git
cd affiliate-marketplace-connector

# CΓ i Δ‘αΊ·t dαΊ‘ng package phΓ‘t triển (Editable Mode)
pip install -e .

After installation, the system registers the following console commands in your Terminal:

  • afflite: All-in-one CLI for multi-platform affiliate management & VideoFactory.

  • affiliate-marketplace-mcp: MCP JSON-RPC 2.0 server over Stdio.

  • shopee-cli: Dedicated CLI for Shopee Research & Cross-Audit.

  • shopee-research: Shopee scanning and extraction toolkit.

  • tiktok-research: TikTok Shop scanning and extraction toolkit.


2. Configure Environment Variables (Optional)

The system automatically uses relative paths within the source directory. To customize, you can set:

# Đường dαΊ«n file SQLite lΖ°u trα»― dα»― liệu
export AFFLITE_DB_PATH="/path/to/your/data/products.db"

# ThΖ° mα»₯c xuαΊ₯t dα»± Γ‘n VideoFactory (AIEV)
export VIDEOFACTORY_PROJECTS_DIR="/path/to/your/VideoFactory/video-projects"

3. Register the MCP Server for AI Assistants

A. For Google Antigravity CLI

Add to Antigravity's MCP config file (~/.gemini/antigravity-cli/mcp_config.json):

{
  "mcpServers": {
    "affiliate-marketplace": {
      "command": "affiliate-marketplace-mcp"
    }
  }
}

B. For Claude Desktop App

Add to ~/Library/Application Support/Claude/claude_desktop_config.json:

{
  "mcpServers": {
    "affiliate-marketplace": {
      "command": "affiliate-marketplace-mcp"
    }
  }
}

C. For Cursor IDE

Add to ~/.cursor/mcp.json or the Cursor Settings > MCP section:

{
  "mcpServers": {
    "affiliate-marketplace": {
      "command": "affiliate-marketplace-mcp"
    }
  }
}

πŸ’» CLI Usage Guide (afflite)

1. Search for Potential Products (search)

# QuΓ©t cαΊ£ 2 sΓ n Shopee & TikTok Shop
afflite search all "chuα»™t cΓ΄ng thΓ‘i học"

# QuΓ©t riΓͺng Shopee
afflite search shopee "bΓ n phΓ­m cΖ‘ khΓ΄ng dΓ’y"

# QuΓ©t riΓͺng TikTok Shop
afflite search tiktok "Δ‘Γ¨n rgb mini quay video"

2. Audit Store Credibility (seller)

# ThαΊ©m Δ‘α»‹nh gian hΓ ng Shopee Mall
afflite seller shopee "Ugreen Official Mall"

# ThαΊ©m Δ‘α»‹nh gian hΓ ng TikTok Shop
afflite seller tiktok "Photoolex Official Store"

3. Extract Product & Variant SKUs (product)

# Tα»± Δ‘α»™ng nhαΊ­n diện Shopee hay TikTok Shop tα»« link hoαΊ·c ID
afflite product "https://shopee.vn/product-url"
afflite product "https://www.tiktok.com/view/product/1729463065317050819"

4. Run the Full Automated Pipeline (pipeline)

# Tα»± Δ‘α»™ng bung tα»« khΓ³a -> Lọc top sαΊ£n phαΊ©m -> ThαΊ©m Δ‘α»‹nh Shop -> BΓ³c tΓ‘ch SKU & Reviews -> LΖ°u DB
afflite pipeline all "phα»₯ kiện setup gΓ³c lΓ m việc"

5. Generate Video Scripts & Integrate VideoFactory (make-video)

# XuαΊ₯t kα»‹ch bαΊ£n 6 cαΊ£nh vΓ  prompt Veo 3.1 ra mΓ n hΓ¬nh terminal
afflite video-prompt "shopee_ugreen_mouse_44426785177"

# TαΊ‘o trọn gΓ³i thΖ° mα»₯c dα»± Γ‘n VideoFactory (AIEV)
afflite make-video "tiktok_prod_1729463065317050819"

# Liệt kΓͺ danh sΓ‘ch tαΊ₯t cαΊ£ cΓ‘c dα»± Γ‘n VideoFactory Δ‘Γ£ sinh
afflite video-projects

6. Reports & Data Export (report / export)

# Xem bΓ‘o cΓ‘o tα»•ng hợp cΓ‘c sαΊ£n phαΊ©m vΓ  biαΊΏn thể trong SQLite
afflite report

# XuαΊ₯t dα»― liệu kα»‹ch bαΊ£n dαΊ‘ng Markdown hoαΊ·c JSON
afflite export "shopee_ugreen_mouse_44426785177" --format md

🐍 Using via Python Code (Programmatic API)

from afflite import (
    run_shopee_search,
    run_shopee_seller,
    run_shopee_product,
    run_tiktok_search,
    run_tiktok_seller,
    run_tiktok_product,
    generate_video_prompt_package,
    get_connection
)
from afflite.video_factory_bridge import create_videofactory_project

# 1. Tìm kiếm sản phẩm TikTok Shop
tiktok_results = run_tiktok_search("Δ‘Γ¨n rgb mini nam chΓ’m")
print(f"TΓ¬m thαΊ₯y {tiktok_results['total_found']} sαΊ£n phαΊ©m tiềm nΔƒng.")

# 2. BΓ³c tΓ‘ch sαΊ£n phαΊ©m & reviews
prod_data = run_tiktok_product("https://www.tiktok.com/view/product/1729463065317050819")
print(f"Đã lΖ°u sαΊ£n phαΊ©m: {prod_data['product_name']} (Điểm Win: {prod_data['win_potential_score']}/10)")

# 3. TαΊ‘o thΖ° mα»₯c dα»± Γ‘n VideoFactory
vf_project = create_videofactory_project(prod_data["product_id"])
print(f"Đã tẑo VideoFactory Project tẑi: {vf_project['project_path']}")
print(f"Xem preview tαΊ‘i: {vf_project['web_ui_url']}")

πŸ—„ SQLite Database Structure (products.db)

The system manages relational data following 3NF normalization standards:

β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”       β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”
β”‚        products         β”‚       β”‚          shops          β”‚
│─────────────────────────│       │─────────────────────────│
β”‚ id (PK)                 β”‚       β”‚ shop_id (PK)            β”‚
β”‚ platform (Shopee/TikTok)│◄──┐   β”‚ shop_name, username     β”‚
β”‚ shop_id (FK) ───────────┼───┼──►│ rating_star, followers  β”‚
β”‚ product_name, brand     β”‚   β”‚   β”‚ response_rate, score    β”‚
β”‚ sale_price_min, max     β”‚   β”‚   β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”¬β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜
β”‚ total_sold, rating_scoreβ”‚   β”‚                β”‚
β”‚ win_potential_score     β”‚   β”‚                β–Ό
β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”¬β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜   β”‚   β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”
             β”‚                β”‚   β”‚      shop_vouchers      β”‚
             β”œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β–Ίβ”‚β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”‚
             β”‚                β”‚   β”‚ voucher_code, discount  β”‚
             β–Ό                β”‚   β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜
β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”   β”‚   β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”
β”‚    product_variants     β”‚   β”‚   β”‚     search_sessions     β”‚
│─────────────────────────│   β”‚   │─────────────────────────│
β”‚ sku_name, color, price  β”‚   β”‚   β”‚ id (PK), seed_idea      β”‚
β”‚ connectivity, status    β”‚   β”‚   β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”¬β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜
β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜   β”‚                β”‚
             β”‚                β”‚                β–Ό
             β–Ό                β”‚   β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”
β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”   β”‚   β”‚     search_results      β”‚
β”‚     product_reviews     β”‚   β”‚   │─────────────────────────│
│─────────────────────────│   β”‚   β”‚ keyword, price_tier     β”‚
β”‚ username, rating        β”‚   β”‚   β”‚ affiliate_win_score     β”‚
β”‚ review_text, media_type β”‚   β”‚   β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜
β”‚ pain_point_tag, score   β”‚   β”‚
β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜   β”‚
             β”‚                β”‚
             β–Ό                β”‚
β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”   β”‚
β”‚  ai_marketing_insights  β”‚   β”‚
│─────────────────────────│   β”‚
β”‚ target_audience, USP    β”‚   β”‚
β”‚ hook_drama, PAS, AIDA   β”‚   β”‚
β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜   β”‚
             β”‚                β”‚
             β–Ό                β”‚
β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”   β”‚
β”‚      product_media      β”‚   β”‚
│─────────────────────────│   β”‚
β”‚ source_url, local_path  β”‚β”€β”€β”€β”˜
β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜

❓ FAQ β€” Frequently Asked Questions

Q1: What mechanism does the affiliate-marketplace-mcp MCP server run on?

The server uses the Model Context Protocol (MCP) standard over Stdio (JSON-RPC 2.0). When AI Assistants (Antigravity, Claude, Cursor) launch, the server automatically establishes a connection via standard input/output without opening any auxiliary network ports, ensuring security and millisecond-level response times.

Q2: How does the SKU variant extraction mechanism in TikTok Shop's "Select options" work?

TikTok Shop PDPs hide variants behind the interactive "Select options" drawer. The system activates a Chrome Automation session to open the drawer, decodes the price structure by each classification (Combo, Color, Power Source), and fully extracts floor/ceiling prices into the product_variants table.

Q3: What does the generated VideoFactory (AIEV) project folder include?

When running afflite make-video <product_id>, the project folder is automatically generated at /Volumes/Agency/VideoFactory/video-projects/<slug> and includes:

  • meta.json: Standard VideoFactory JSON format with 6 CRO scenes (Hook, Problem, Reveal, USP, Social Proof, CTA).

  • hyperframes.json & index.html: HTML5/CSS3 HyperFrames motion graphics structure at 1080x1920 30fps.

  • assets/hero.jpg: Sharp original product image.

  • assets/transcript_tts.txt: Vietnamese voice-over transcript with per-second timestamps for use with VieNeu-TTS / OmniVoice.

Q4: How do I change the SQLite products.db file when running on another machine?

Simply set the environment variable:

export AFFLITE_DB_PATH="/duong_dan_moi/products.db"

All CLI tools and the MCP Server will automatically point to the new path without any code changes.

Q5: What formulas are used to calculate the Affiliate Win Score and Seller Health Score?

  • Affiliate Win Score (0 - 10): Calculated based on 4 weighted factors: Verified sales count (35%), Star rating score (25%), Impulse Buy fast-conversion price range 150k-500k (20%), and Shopee Mall / Official Shop store status (20%).

  • Seller Health Score (0 - 100): Assessed based on: Average star rating (35 pts), On-time delivery rate (25 pts), Low cancellation rate under 1% (20 pts), Chat response speed & follower count (20 pts).

Q6: Is there support for video generation commands for Google Flow (Veo 3.1)?

Yes. The 6-scene scripts in video-prompt and make-video are specially designed with detailed visual prompts (Extreme close-up, 3D panning, 8k photorealistic, lighting setup) that are 100% compatible with Google Flow (gflow CLI Veo 3.1).


πŸ§ͺ Automated Testing (Unit Testing)

Run the test suite with 18 test scenarios:

python3 -m unittest discover tests

Results:

..................
----------------------------------------------------------------------
Ran 18 tests in 4.228s

OK

The project is distributed under the MIT License. Developed and maintained by tody-agent.

Install Server
A
license - permissive license
A
quality
B
maintenance

Maintenance

–Maintainers
–Response time
–Release cycle
–Releases (12mo)
Commit activity

Resources

Unclaimed servers have limited discoverability.

Looking for Admin?

If you are the server author, to access and configure the admin panel.

Related MCP Servers

  • F
    license
    Not graded
    quality
    D
    maintenance
    Connects AI agents to the Shopee platform for real-time product searches and category mapping via the Affiliate and Seller Center APIs. It enables users to filter products by keyword, category, and commission rates directly through natural language.
    1
  • A
    license
    D
    quality
    D
    maintenance
    Enables natural-language control of e-commerce operations including product management, order processing, inventory tracking, customer service, content generation, and advertising analytics through 15 integrated MCP tools. Provides a local-first commerce automation solution with SQLite storage and extensible channel adapters for end-to-end online store workflows.
    15
    3
    MIT

View all related MCP servers

Related MCP Connectors

  • Hosted MCP with 91 agent tools: X, domains, SEO, Maps, Trends, Search, YouTube, TikTok, and more.

  • 100+ MCP tools for AI agents: content metadata, trade intelligence, business-expertise analysis.

  • Free public MCP for AI agents β€” 193 tools, 44 workflows. No API key.

View all MCP Connectors

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/tody-agent/affiliate-marketplace-connector'

If you have feedback or need assistance with the MCP directory API, please join our Discord server