Quickstart
Connect Pangolinfo MCP to your AI Agent in 3 minutes — remote HTTP transport, zero install, zero dependencies, always tracks the latest version.
This page covers the general Streamable HTTP setup steps. For per-client config locations, see Client Setup.
Connection Info
| Field | Value |
|---|---|
| Server URL | https://mcp.pangolinfo.com/mcp |
| Transport | Streamable HTTP |
| MCP Protocol | Compatible with MCP standard (version auto-negotiated by client) |
| Auth | Header Authorization: Bearer <YOUR_API_KEY> |
| API Key | Pangolinfo Console → Profile → API Key |
Configuration Tip: Ensure your MCP uses a Permanent Key; Short-Term Keys will expire and silently break your AI workflow.
3 Steps to Connect
Get a Permanent API Key
Log into the Pangolinfo Console → go to Profile → API Key, copy your Permanent Key. Use a Permanent Key for MCP deployment to prevent authentication expiry.
Add URL + Key to your client MCP config
Open your AI client’s MCP config file (each client uses a different path — see “Universal Config Template” below) and add a Pangolinfo entry.
Restart the client and wait for the tool list to load
After restarting the AI client, you should see 19 tools in the MCP tools panel. On your first conversation, ask the AI to call pangolinfo_capabilities to introspect.
Universal Config Templat
Most clients follow a similar JSON structure. Replace <YOUR_API_KEY> with your Permanent Key:
Field names vary slightly across clients (e.g. type: "http", transport: "streamable-http"). See Client Setup for the exact path and fields each client expects.
Verify Connection
After restarting the AI client, send it:
If you get back 19 tools with brief descriptions, you’re connected. If you see “no tools” or “auth failed”, check FAQ.
Try Your First Query
Once connected, you can directly say to the AI:
“Pull product details for ASIN B0DYTF8L2W and summarize the top 3 positive and top 3 negative review points.”
The AI will call get_amazon_product (amzProduct parser) and read the aiReviewsSummary field from the response — one tool call covers it, no need to paginate raw reviews.
The product detail endpoint (amzProduct parser) returns an aiReviewsSummary (AI review summary) field. Most “summarize reviews” requests can be done in a single call, avoiding the extra credit cost of paginated review scraping. Use get_amazon_reviews only when you need individual raw reviews.
For more pre-built workflows, see Tool Reference → 6 Preset Workflows.

