Introduction
Pangolinfo MCP for Agent connects AI agents to structured market data and consumer intelligence through remote MCP servers.
MCP for Agent
Pangolinfo MCP for Agent lets AI clients call Pangolinfo data and insight tools through the Model Context Protocol. Instead of switching between dashboards, scripts, and spreadsheets, you can ask an AI agent to collect data, read structured outputs, compare signals, and generate analysis in the same conversation.
MCP (Model Context Protocol) is an open protocol released by Anthropic that lets AI agents call external tools through a unified interface. Pangolinfo MCP wraps data access, market analysis, VOC intelligence, and risk-screening capabilities into MCP servers, so AI agents can work from real-time data instead of model memory alone.
For AI Agent Users
Add an MCP URL and Pangolinfo API Key to Claude Code, Cursor, Cline, or another supported client, and the AI can call data and insight tools directly in the conversation.
For Developers
MCP calls Pangolinfo data services and business APIs underneath. MCP is the semantic layer for AI; REST API remains the better fit for scripts, system integration, and batch jobs.
MCP server configuration too complex? Non-technical users can also use tool-type SKILLs, which provide similar capabilities with a lighter setup path.
Pangolinfo currently provides two MCP services:
Amazon Insight MCP
Turn structured market data into actionable insights.
Access real-time Amazon, search engine, and patent data through unified AI-powered tools.
VOC Insight MCP
Turn consumer conversations into actionable insights.
Analyze multi-platform discussions to generate AI-powered VOC, sentiment, and competitor intelligence.
How it Works
- Choose the MCP service that matches your workflow.
- Connect the MCP URL and your Pangolinfo API Key in a supported AI client.
- Ask the AI agent a business question in natural language.
- The agent selects the right tools, calls Pangolinfo, and returns structured analysis.
When to Use MCP
| Scenario | MCP | REST API |
|---|---|---|
| Get data inside an AI conversation | Let the AI call tools directly without switching context | Write scripts or send manual requests |
| Chain multiple tools | The AI selects tools and combines results | You build the orchestration logic |
| Discover required parameters | The AI reads tool schemas | You read API documentation |
| Manage credentials | Configure one URL and API Key in the client | Handle auth in every project or script |
| Batch automation | Best for lightweight automation and interactive analysis | Best for large-scale batch jobs, CI, and system integration |
Rule of thumb: use MCP for conversational research, product analysis, VOC, and report generation; use REST API for fixed large-scale batch workflows.
Supported Clients
Amazon Insight MCP and VOC Insight MCP both use remote Streamable HTTP and work with mainstream AI Agent clients such as Claude Code, Cursor, Cline, Windsurf, Codex, Hermes, OpenClaw, and WorkBuddy. Client config fields vary slightly; see Client Setup for templates.
Amazon Insight MCP Tools at a Glance
| Category | Representative Capabilities |
|---|---|
| Amazon core data | search_amazon / get_amazon_product / get_amazon_reviews / list_seller_products / list_bestsellers / list_new_releases / scrape_url |
| Amazon categories and niche analytics | get_category_children / search_categories / get_category_paths / list_category_products / filter_categories / filter_niches |
| Search and SERP AI | ai_search / keyword_trends / search_amazon_alexa |
| Maps and places | search_local_maps |
| Design patent and litigation compliance | wipo_search, with optional US patent-litigation chaining |
| MCP introspection | pangolinfo_capabilities |
For full field explanations and field dependencies, see Tool Reference. For common usage chains, see Amazon Insight MCP Workflow.
VOC Insight MCP Capabilities at a Glance
| Category | Representative Capabilities |
|---|---|
| Context and rules | Read account context, brand list, supported platforms, product rules, billing rules, and suggested next actions |
| Knowledge-space onboarding | Generate a collection plan from a brand, product, or topic; confirm industry, keywords, platforms, pages, and estimated points before creating a space |
| Brand management | View brand configuration, prepare full onboarding, and update keywords, platforms, competitors, and brand description |
| Collection refresh | Diagnose freshness, start refresh, poll collection progress, and avoid duplicate refreshes while collection is running |
| Data reading | Read brand metrics, search posts, semantic-search posts, sentiment, voice share, competitor comparison, and risk alerts |
| AI analysis | Generate AI deep-analysis reports after collection completes, or read a quick brand summary |
Default coverage includes TikTok, Instagram, YouTube, X, Facebook, Pinterest, and Trustpilot. Threads and Reddit can be added when needed. Reddit counts as 2 weighted channel units. For VOC formulas, tool IDs, error codes, and report interpretation, see:
Which MCP Should I Use?
| Need | Recommended MCP |
|---|---|
| Amazon product, keyword, review, seller, category, search, or patent analysis | Amazon Insight MCP |
| Brand social listening, VOC, sentiment, competitor comparison, risk monitoring | VOC Insight MCP |
| A mixed market research workflow | Connect both MCP services and let the agent route tasks |