# Pangolinfo Amazon Reviews MCP An official, read-only MCP server focused on Amazon review research and voice-of-customer (VOC) signals. It gives AI agents a smaller, review-specialized tool surface while using the same authenticated Pangolinfo data engine as the broader [Amazon Data MCP](https://www.pangolinfo.com/amazon-data-mcp/). ## Tools - `get_amazon_reviews` — paginated raw reviews - `find_complaint_signals` — low-rating complaint evidence - `summarize_review_ratings` — deterministic rating counts and averages - `filter_review_media` — image/video evidence reviews - `compare_review_pages` — side-by-side rating and complaint metrics for two ASINs All tools are read-only. The bridge never stores API keys or result data. It forwards requests to the hosted MCP endpoint and performs deterministic local filtering only. ## Quick start Get a Pangolinfo key from the [Amazon Review API page](https://www.pangolinfo.com/amazon-review-api/), then configure any MCP client: ```json { "mcpServers": { "pangolinfo-amazon-reviews": { "command": "npx", "args": ["-y", "pangolinfo-amazon-reviews-mcp"], "env": { "PANGOLINFO_API_KEY": "" } } } } ``` Or run `PANGOLINFO_API_KEY=... npx -y pangolinfo-amazon-reviews-mcp`. Docker and hosted HTTP setup are documented on the [Amazon Review API landing page](https://www.pangolinfo.com/amazon-review-api/) and [Pangolinfo docs](https://docs.pangolinfo.com/). ## Links - Product: https://www.pangolinfo.com/amazon-review-api/ - Docs: https://docs.pangolinfo.com/ - Source: https://github.com/Pangolin-spg/amazon-reviews-mcp - Amazon Data MCP: https://www.pangolinfo.com/amazon-data-mcp/ ## License MIT. Pangolinfo trademarks and logos remain subject to the brand guidelines in `BRANDING.md`.