Skip to main content
Glama
Pangolin-spg

Pangolinfo Amazon Data MCP

README.md
# pangolinfo-mcp

[![Listed on mcpservers.org](https://mcpservers.org/badge.svg)](https://mcpservers.org/servers/pangolin-spg/pangolinfo-mcp)

Python client for the **Pangolinfo Amazon Data MCP** server — 19 Amazon / WIPO / PACER data tools callable from Python, no AI client required.

> Built and maintained by [PANGOLIN INFO TECH PTE. LTD.](https://www.pangolinfo.com)

## Why this package

The [Amazon Data MCP](https://www.pangolinfo.com/amazon-data-mcp/) gives AI agents 19 data tools via the MCP protocol. This package lets **Python developers** call those same tools from code — no Claude, Cursor, or any AI client needed.

- **19 tools** in one client: Amazon product, reviews, search, niches, bestsellers, AI SERP, WIPO, PACER and more
- **Zero dependencies** beyond `httpx` — no MCP SDK required
- **Sync** interface (async coming soon)
- **Same engine, same pricing** as the REST API and MCP server

## Installation

```
pip install pangolinfo-mcp
```

## Quick start

### 1. Get a permanent API key

Sign up at [tool.pangolinfo.com](https://tool.pangolinfo.com) and grab a permanent API key from Account Center.

### 2. Use the client

```python
from pangolinfo_mcp import PangolinfoMCPClient

client = PangolinfoMCPClient(api_key="your_permanent_api_key")

# Amazon product detail
product = client.get_amazon_product("B0DYTF8L2W")
print(product["title"])

# Keyword search
results = client.search_amazon("wireless earbuds")

# Best sellers in a category
bestsellers = client.list_bestsellers(category_id="172282")

# WIPO trademark check
trademarks = client.wipo_search("Nike")

# PACER patent litigation
cases = client.pacer_search("Apple Inc.")

# List all available tools
tools = client.list_tools()

client.close()
```

### Context manager

```python
with PangolinfoMCPClient(api_key="your_key") as client:
    product = client.get_amazon_product("B0DYTF8L2W")
```

## All 19 tools

### Amazon Core Data (5)
| Method | Description |
|---|---|
| `client.search_amazon(keyword)` | Search Amazon products |
| `client.get_amazon_product(asin)` | Full product detail |
| `client.get_amazon_reviews(asin)` | Paginated reviews |
| `client.list_seller_products(seller_id)` | Seller storefront |
| `client.scrape_url(url)` | Scrape any Amazon page |

### Category & Niche Analysis (8)
| Method | Description |
|---|---|
| `client.filter_niches(**metrics)` | Blue-ocean niche discovery |
| `client.filter_categories(**metrics)` | Category filtering |
| `client.search_categories(query)` | Full-text category search |
| `client.get_category_children(node_id)` | Browse node children |
| `client.get_category_tree()` | Root category tree |
| `client.batch_category_paths(category_ids)` | Batch category paths |
| `client.list_bestsellers(category_id)` | Best Sellers list |
| `client.list_new_releases(category_id)` | New Releases list |

### Search & SERP AI (3)
| Method | Description |
|---|---|
| `client.ai_search(query)` | AI Overview / AI Mode SERP |
| `client.keyword_trends(keyword)` | Keyword trend time-series |
| `client.search_amazon_alexa(query)` | Alexa for Shopping |

### Maps & IP Compliance (3)
| Method | Description |
|---|---|
| `client.search_local_maps(query)` | Maps POI data |
| `client.wipo_search(keyword)` | WIPO trademark search |
| `client.pacer_search(query)` | PACER patent litigation |

### Utility (1)
| Method | Description |
|---|---|
| `client.capabilities()` | List all live tools (free) |

## Generic tool call

```python
# Call any tool by name
result = client.call_tool("search_amazon", {"keyword": "laptop", "page": 2})
```

## Use as an MCP server

This repo also documents the **hosted MCP server** that any MCP-compatible client (Claude Desktop, Cursor, VS Code, etc.) can connect to directly over Streamable HTTP — no local install required.

Add to your client's `mcp.json` / MCP config:

```json
{
  "mcpServers": {
    "pangolinfo": {
      "url": "https://mcp.pangolinfo.com/mcp",
      "headers": {
        "Authorization": "Bearer pgl_your_api_key"
      }
    }
  }
}
```

Or pass the key inline as a query parameter:

```
https://mcp.pangolinfo.com/mcp?api_key=pgl_your_api_key
```

Get a permanent API key at [tool.pangolinfo.com](https://tool.pangolinfo.com). The key is sent as a Bearer token; config is not hot-reloaded, so restart/reconnect the client after changing it.

For MCP clients that require a local stdio process, use the official npm bridge:

```bash
PANGOLINFO_API_KEY="your_key" npx -y pangolinfo-amazon-data-mcp
```

Source and client configuration: [github.com/Pangolin-spg/amazon-data-mcp](https://github.com/Pangolin-spg/amazon-data-mcp).


## Links

- **MCP product page:** [pangolinfo.com/amazon-data-mcp](https://www.pangolinfo.com/amazon-data-mcp/)
- **Documentation:** [docs.pangolinfo.com](https://docs.pangolinfo.com)
- **Get a free API key:** [tool.pangolinfo.com](https://tool.pangolinfo.com)
- **Source code:** [github.com/Pangolin-spg/pangolinfo-mcp](https://github.com/Pangolin-spg/pangolinfo-mcp)

## License

MIT — © PANGOLIN INFO TECH PTE. LTD.

TDQS

A4.2/5.0

Scored across 21 tools

Disambiguation5/5

Nearly every tool carries explicit 'Use when' / 'Don't use' guidance that names the sibling tool to prefer instead (e.g. get_amazon_product vs get_amazon_delivery_time vs get_amazon_reviews; list_bestsellers vs list_new_releases vs list_category_products). The scrape_url escape hatch and the four search tools (search_amazon, ai_search, search_amazon_alexa, search_local_maps) are each bounded by source and intent, so an agent can reliably pick one.

Naming Consistency4/5

Most names follow a clean verb_noun pattern (search_amazon, get_amazon_product, list_bestsellers, filter_niches, get_category_paths). A few deviate: ai_search and wipo_search invert the order, keyword_trends is noun_noun, and pangolinfo_capabilities uses a vendor prefix, but all remain readable and predictable.

Tool Count4/5

21 tools is on the heavy side but justified by the breadth of the platform (Amazon product/review/category/niche data plus Google, Maps, WIPO, and Alexa sources). Each tool covers a distinct surface, though a couple (search_amazon_alexa, get_amazon_alexa_questions) feel like niche additions that push the count near the practical ceiling.

Completeness4/5

The surface covers the full data-retrieval lifecycle: keyword search, product/PDP detail, reviews, category tree drilling, category/niche commercial-metric filtering, seller storefronts, and external demand/IP signals, with a generic scrape_url fallback. Account/credits are deliberately excluded and there is no keyword-rank-history tool, but core workflows have no dead ends.

Maintenance

ActivityMaintained
ResponsivenessNo issues