pangolinfo-mcp
Health Warn
- License — License: MIT
- Description — Repository has a description
- Active repo — Last push 0 days ago
- Low visibility — Only 6 GitHub stars
Code Warn
- process.env — Environment variable access in src/auth.ts
- process.env — Environment variable access in src/i18n.ts
Permissions Pass
- Permissions — No dangerous permissions requested
No AI report is available for this listing yet.
Pangolinfo MCP server — Amazon e-commerce data tools for AI assistants via MCP protocol
Pangolinfo MCP
One MCP service for synchronous Amazon research, ecommerce Product and Reviews, AI search, keyword trends and profit calculations.
- Version: 0.10.0
- Tools: 21 total: 20 business tools and free local
pangolinfo_capabilities. - Hosted endpoint:
https://mcp.pangolinfo.com/mcp(Streamable HTTP). - Authentication: your Pangolinfo API key; existing endpoint and client configuration remain valid.
- Scope: seven ecommerce platforms, existing Amazon/category/niche tools and V1/V2 synchronous APIs. Social media, all V3 APIs and asynchronous jobs/callbacks/polling are excluded.
Upgrade from 0.9.x
collect_social_data, search_naver_cafe, search_local_maps and wipo_search are no longer registered. Refresh your client's tool cache. The independent VOC Insight MCP is retired; its social/brand/report/persona tools are not migrated here. REST services and stored data are unchanged.
calculate_amazon_profit is new. collect_ecommerce_data exposes seven platforms with product-detail and reviews. Use Product for combined details and aggregate ratings; Shopify and the legacy product-summary operation are outside this MCP catalog.
Install
Recommended: one-line installer (covers 7 AI clients)
The Pangolinfo Installer detects your AI client, writes the right config files, and you're done. Pass --scope=mcp to install only this MCP server (skip the Skills package).
macOS / Linux
curl -fsSL https://pangolinfo.dev/install.sh | sh -s -- \
--agent=<your-agent> \
--scope=mcp \
--api-key=pgl_xxxxxxxxxxxx
Windows (PowerShell)
irm https://pangolinfo.dev/install.ps1 | iex; `
Install-Pangolinfo -Agent <your-agent> -Scope mcp -ApiKey pgl_xxxxxxxxxxxx
<your-agent> is one of: claude-code, cursor, cline, windsurf, codex, hermes, openclaw.
After the installer finishes, restart your AI client so it picks up the new mcpServers entry.
Manual install (just download one file)
The release artifact is a single self-contained server.mjs (~800 KB) — all dependencies are bundled in. No npm install needed.
# macOS / Linux
mkdir -p ~/.local/lib/pangolinfo-mcp
curl -fsSL https://github.com/pangolinfo/pangolinfo-mcp/releases/latest/download/server.mjs \
-o ~/.local/lib/pangolinfo-mcp/server.mjs
chmod +x ~/.local/lib/pangolinfo-mcp/server.mjs
# Windows (PowerShell)
$dir = "$env:LOCALAPPDATA\pangolinfo-mcp"
New-Item -ItemType Directory -Force -Path $dir | Out-Null
irm https://github.com/pangolinfo/pangolinfo-mcp/releases/latest/download/server.mjs `
-OutFile "$dir\server.mjs"
Then wire it into your AI client — see the per-client snippets below. Point args at the file you just downloaded.
Developers: to build from source,
git clonethis repo and runnpm install && npm run build. The produceddist/server.mjsis identical to the release asset.
Get an API key
- Sign up at https://tool.pangolinfo.com/#/en/system/loading?sourceTag=github_amz
- Copy your
pgl_xxxxxxxxkey from the dashboard - Top up credits if needed (charges depend on the data tool and format;
pangolinfo_capabilitiesis free)
Manual configuration (per AI client)
Replace /abs/path/to/pangolinfo-mcp/dist/server.mjs with your real path, and pgl_xxxxxxxx with your key.
Claude Code (~/.claude/settings.json)
{
"mcpServers": {
"pangolinfo": {
"command": "node",
"args": ["/abs/path/to/pangolinfo-mcp/dist/server.mjs"],
"env": { "PANGOLINFO_API_KEY": "pgl_xxxxxxxx" }
}
}
}
Prefer
claude mcp add --scope user pangolinfo node /abs/path/to/dist/server.mjs— it writes the same entry without hand-editing JSON.
Cursor (~/.cursor/mcp.json)
{
"mcpServers": {
"pangolinfo": {
"command": "node",
"args": ["/abs/path/to/pangolinfo-mcp/dist/server.mjs"],
"env": { "PANGOLINFO_API_KEY": "pgl_xxxxxxxx" }
}
}
}
Cline (VS Code extension)
Open Cline → MCP Servers → Edit settings JSON, then add:
{
"mcpServers": {
"pangolinfo": {
"command": "node",
"args": ["/abs/path/to/pangolinfo-mcp/dist/server.mjs"],
"env": { "PANGOLINFO_API_KEY": "pgl_xxxxxxxx" }
}
}
}
The settings file lives at <vscode-user>/globalStorage/saoudrizwan.claude-dev/settings/cline_mcp_settings.json.
If you're on the standalone Cline CLI, use ~/.cline/data/settings/cline_mcp_settings.json instead.
Windsurf (~/.codeium/windsurf/mcp_config.json)
{
"mcpServers": {
"pangolinfo": {
"command": "node",
"args": ["/abs/path/to/pangolinfo-mcp/dist/server.mjs"],
"env": { "PANGOLINFO_API_KEY": "pgl_xxxxxxxx" }
}
}
}
Codex (~/.codex/config.toml)
[mcp_servers.pangolinfo]
command = "node"
args = ["/abs/path/to/pangolinfo-mcp/dist/server.mjs"]
[mcp_servers.pangolinfo.env]
PANGOLINFO_API_KEY = "pgl_xxxxxxxx"
Hermes (~/.hermes/config.yaml)
mcp_servers:
pangolinfo:
command: node
args: ["/abs/path/to/pangolinfo-mcp/dist/server.mjs"]
env:
PANGOLINFO_API_KEY: pgl_xxxxxxxx
OpenClaw (~/.openclaw/openclaw.json)
{
"mcpServers": {
"pangolinfo": {
"command": "node",
"args": ["/abs/path/to/pangolinfo-mcp/dist/server.mjs"],
"env": { "PANGOLINFO_API_KEY": "pgl_xxxxxxxx" }
}
}
}
Hosted (remote HTTP) — no local install
Don't want to install anything? Point your AI client at the hosted endpoint
instead of a local server.mjs. This is the streamable HTTP transport
(MCP spec), multi-tenant — you bring your own key on every request.
| Endpoint | https://mcp.pangolinfo.com/mcp |
| Transport | Streamable HTTP (POST + SSE) |
| Auth | Your pgl_xxxxxxxx key, via Bearer header or URL query |
Passing your API key — two ways
1. Authorization header (recommended)
Authorization: Bearer pgl_xxxxxxxx
This is the preferred method: the key stays out of URLs, logs, and browser
history. The Bearer scheme is matched case-insensitively (bearer,BEARER, Bearer all work) and extra whitespace is tolerated.
2. ?api_key= URL query (fallback)
https://mcp.pangolinfo.com/mcp?api_key=pgl_xxxxxxxx
Easiest for clients that only let you paste a URL. If your key contains
URL-special characters, URL-encode it. The header (method 1) takes
precedence if both are supplied.
Client config example (Claude Code / Cursor — "url"-style remote MCP)
{
"mcpServers": {
"pangolinfo": {
"url": "https://mcp.pangolinfo.com/mcp",
"headers": { "Authorization": "Bearer pgl_xxxxxxxx" }
}
}
}
If your client can't set headers, fall back to the URL form:
"url": "https://mcp.pangolinfo.com/mcp?api_key=pgl_xxxxxxxx".
Notes for custom / raw HTTP integrations
If you're calling the endpoint directly (not through a standard MCP client):
- POST your JSON-RPC body to
/mcp. - The MCP streamable transport normally requires the request to advertise
bothapplication/jsonandtext/event-streamin itsAcceptheader.
This server backfills the missing media type for you, so a plainAccept: application/json,Accept: */*, or even a missingAccept
header all work — you don't need to hand-craft it. - A
401means the key was not found in either the header or the
query string — check that your client actually sent it (some libraries
drop theAuthorizationheader on cross-origin redirects).
Tools (21 total)
| Tool | Capability | Points |
|---|---|---|
pangolinfo_capabilities |
Current catalog and workflows; no backend call | Free |
search_amazon |
Amazon keyword search | JSON 1/page; rawHtml/markdown 0.75/page |
get_amazon_product |
Amazon product details | JSON 1/page; rawHtml/markdown 0.75/page |
get_amazon_delivery_time |
Product details plus delivery estimates | 2 |
get_amazon_reviews |
Amazon buyer reviews | 5/page |
list_bestsellers |
Best sellers | JSON 1/page; rawHtml/markdown 0.75/page |
list_new_releases |
New releases | JSON 1/page; rawHtml/markdown 0.75/page |
list_seller_products |
Seller listings | JSON 1/page; rawHtml/markdown 0.75/page |
list_category_products |
Category listings | JSON 1/page; rawHtml/markdown 0.75/page |
search_categories |
Category keyword search | 2/call |
get_category_children |
Category children | 2/call |
filter_categories |
Category analytics | 5/call |
filter_niches |
Niche analytics | 10/call |
get_category_paths |
Category breadcrumb paths | 2/call |
ai_search |
AI Overview / AI Mode | 2 |
keyword_trends |
Keyword trends | 1.5 |
scrape_url |
Generic Amazon pages; includes competing offers | Format/parser-specific; offers 1 |
search_amazon_alexa |
Alexa Agent answers | 6/answered turn |
get_amazon_alexa_questions |
Alexa Listing questions and answers | 0/5 discovery + 20/answered region, max 45 |
collect_ecommerce_data |
Seven-platform Product and Reviews | 0.8/product or unique review |
calculate_amazon_profit |
US/JP FBA and merchant-fulfilled fees, profit and margin | 5 |
Billing follows the corresponding REST API rules. Preserve pointCost when returned; an absent field does not mean zero cost. Revenue Calculator does not return pointCost and costs a fixed 5 points per successful call. Unavailable-ASIN and redirect outcomes have their own billing rules.
Profit example
{"asin":"B07FZ8S74R","country":"US"}
country accepts US/JP, case-insensitively. Optional prices, shipping, package dimensions/weight and programs follow the Revenue Calculator API. Returns code, message and data containing summary, product and raw; there is no pointCost field. Product-not-found code 8000 is unbilled and maps to PRODUCT_NOT_FOUND, retriable=no; check ASIN and marketplace before changing the request.
Profit calculation has a 180-second local wait budget; allow at least 210 seconds in your MCP client. This budget is not a bound on backend retries and cannot override a gateway timeout. A timeout, connection failure or HTTP 5xx maps to OUTCOME_UNKNOWN, retriable=no: completion and billing are unknown. Check call/credit records before deciding to submit again. A confirmed business failure keeps its normal classification. Before release, verify both the client-to-MCP and MCP-to-REST gateway limits; increasing this local budget alone cannot guarantee delivery or backend cancellation.
Ecommerce example
{"platform":"walmart","operation":"product-detail","country":"US","productId":"103158059"}
Product combines details and aggregate ratings; Reviews returns buyer reviews. See ecommerce parameters and the MCP tool reference.
For competing offers, call scrape_url with parserName=amzFollowSeller and either an ASIN in content or an Amazon product URL. Each returned item includes soldBy, hasSoldByLink (linked seller name vs. plain text), and optional isFeatured for the featured offer.
Review code 9206 confirms an ASIN redirect (for example B0GGC6B8ZD → B000XH82O6). MCP reports ASIN_REDIRECTED, retriable=no, preserving the destination message and pointCost. The original review request is billed normally. Empty review pages may trigger an additional product-page verification, which adds latency but has no separate charge. Do not automatically retry or fetch the destination ASIN.
Follow Seller code 9206 means a confirmed product redirect and is billed as a normal request (usually 1 credit; see pointCost). message contains the destination, for example Redirected to ASIN B0DSHTVHNZ. data.json retains the empty result for the original request; the destination product is not queried automatically. Do not automatically retry the original ASIN. Query the destination separately only after confirming it; that request is billed separately. Fetch failures, mismatched delivery destinations, and unconfirmed redirects remain 9200 without a charge. MCP reports [ASIN_REDIRECTED] with retriable=no and preserves the backend message and pointCost in error details.
When the product-detail check explicitly confirms that the requested ASIN is missing on the current marketplace, Follow Seller returns 9204 with message: ASIN not found on this marketplace: B08HSS37H7 (the requested ASIN). This is billed normally, usually 1 credit; see pointCost. Do not automatically retry the same marketplace and ASIN; check the marketplace and identifier first. The ASIN may still exist on another marketplace. An empty or 404 offers page alone, network failures, challenges, or inconclusive evidence do not establish that the ASIN is missing; unconfirmed results remain 9200 without a charge. MCP reports [ASIN_UNAVAILABLE] with retriable=no and preserves the message and credit cost.
search_amazon_alexa returns a status on every turn: answered, refused (Amazon declined the question; content and products are empty and refusal_message holds Amazon's wording, e.g. その質問にはお答えできません。) or failed. Only answered turns are billed (6 credits each). When Amazon declines every turn the call fails with ALEXA_REFUSED (code 9207, not billed, retriable=no); the same prompt will be declined again, so rephrase it rather than retrying.
US search_amazon_alexa queries optionally accept asin or a product url; both must identify the same product when supplied together. The response includes asin and context_source: "product_page" only after the real Amazon PDP request context is verified. The product panel is used without requiring inline preset-question buttons. get_amazon_alexa_questions falls back to initial side-panel suggestions when both inline regions are empty; these appear in regions.panel with source: "panel". With region: "all", panel answers count as one answer region (25 credits including discovery); no available suggestions costs zero.
For product subtitles/highlights, read get_amazon_product JSON at data.json[0].data.results[0].itemHighlights. The parser normalizes the dedicated subtitle, the main-title suffix, or an already parsed Item Highlight(s) attribute into this string; it is empty when unavailable. Markdown conversion is not required.
Plus a free local call:
pangolinfo_capabilitiesreturns the full tool catalog, canonical workflows, and usage tips with no backend round-trip (0 credits). It is a self-introspection helper, one of the 21 registered tools, alongside 20 business tools.
Default marketplace is Amazon US (marketplaceId=ATVPDKIKX0DER, zip=90001). Override per call via tool arguments.
⏱️ Slow tools & your client's tool-call timeout
Two tools legitimately run long because they wait on live AI generation:
| Tool | Typical latency | Worst case |
|---|---|---|
search_amazon_alexa (Rufus) |
60–90s per prompt | >200s for multiple prompts |
ai_search |
~30s | ~60s |
Most MCP clients enforce a per-tool-call timeout — often 60 seconds of silence — and abort the call if nothing comes back in time. When that happens the client reports a transport-level timeout/disconnect, and the agent mistakes it for "the tool is unavailable/broken." This is the usual reason search_amazon_alexa gets flagged as unavailable — it reliably crosses the silent 60s line.
Two things keep this from happening:
The server emits progress heartbeats. While a tool runs, the server sends a
notifications/progressevery 15s if your client included aprogressTokenin the call. Spec-compliant clients reset their idle timeout on each heartbeat, so the whole 60–90s render stays under the wire. Most modern MCP clients send aprogressTokenautomatically — no action needed.Raise the client timeout for clients that don't honor progress. If your client has no
progressTokensupport or a hard cap, bump its MCP tool-call timeout to ≥120s before usingsearch_amazon_alexa. Where to set it depends on the client (examples):- Claude Code / config-based clients: raise the MCP request/tool timeout in the client config.
- Custom SDK clients: pass a larger
timeout(and, ideally, aprogressToken) in thecallToolrequest options.
Also prefer exactly one prompt per search_amazon_alexa call — multiple prompts stack latency linearly and make the timeout far more likely.
Auth resolution order
The server resolves the API key with this priority:
- CLI args:
--api-key=pgl_xxx --api-base=... --scrape-base=... - Env vars:
PANGOLINFO_API_KEY,PANGOLINFO_API_BASE,PANGOLINFO_SCRAPE_BASE - Config file at
~/.pangolinfo/config.json:{ "api_key": "pgl_xxxxxxxxxxxx", "api_base": "https://extapi.pangolinfo.com", "scrape_base": "https://scrapeapi.pangolinfo.com" } - Missing key → startup failure with an actionable error.
CLI args win over env vars — convenient when you want per-server keys without polluting the global environment.
Internationalization
Tool descriptions and error hints are available in English and Chinese. The
language is resolved in this order: --lang=zh|en → PANGOLINFO_LANG=zh|en → OS locale
($LANG starting with zh* → Chinese, otherwise English) → English when there is no
locale signal at all. So a Chinese-locale machine gets Chinese automatically; everyone
else gets English. Force a language explicitly:
"env": {
"PANGOLINFO_API_KEY": "pgl_xxxxxxxx",
"PANGOLINFO_LANG": "zh"
}
Startup logs are always English (operator-facing); tool descriptions and error hint fields follow the resolved locale.
Verify your install
After restarting your AI client, ask it:
List all available
pangolinfoMCP tools.
You should see 21 tools (20 business tools plus free introspection). Then try:
Use
pangolinfo_capabilitieswith mode "summary".
This is a free local call — if it returns the tool catalog, your install is wired correctly. Next, run something paid like:
Search Amazon for "wireless mouse" and return the top 5 results.
Expected: ~0.75 credits deducted, ~300 KB of structured product data returned.
Development
npm install
npm run dev # tsx src/server.ts — hot-reload
npm run build # esbuild → dist/server.mjs
npm run typecheck # tsc --noEmit
npm start # node dist/server.mjs
Project layout
src/
├── server.ts MCP stdio entry + tool registration
├── auth.ts API key resolution (CLI > env > config file)
├── client.ts HTTP client (Authorization, User-Agent)
├── errors.ts PangolinfoError + status-code mapping
├── config.ts Default endpoints / constants
├── i18n.ts zh/en translation lookup
└── tools/
├── _types.ts Tool / ToolContext type definitions
├── index.ts Tool registry (21 total: 20 business + capabilities)
└── <verb_noun>.ts One file per tool
Adding a new tool
- Create
src/tools/<verb_noun>.tsexporting aToolobject — mirrorsearch_amazon.ts. - Import it in
src/tools/index.tsand append to thetoolsarray. - Schema is
zod;.describe()every field — the AI reads those. - Never call
fetchdirectly — usectx.client.post(...). Auth is already injected. - Throw
PangolinfoErroron failure; the HTTP client already throws this for non-2xx responses.
Security & Data Handling
We take operator and user safety seriously. By design, this MCP server:
- Brings your own key. Authentication is via your personal
PANGOLINFO_API_KEY(issued at https://tool.pangolinfo.com/#/en/system/loading?sourceTag=github_amz). The key is read locally from your AI client's config or environment — it is never transmitted anywhere except tohttps://scrapeapi.pangolinfo.com(orhttps://mcp.pangolinfo.comfor the hosted variant) over TLS 1.2+. - No telemetry. This server does not phone home, does not collect usage analytics, and does not log your prompts. The only outbound traffic is the Amazon / ecommerce / search data API calls you explicitly invoke through tools.
- No PII collection. No user account info, no email, no IP geolocation, and no prompt content is persisted by this server. Tool calls forward only the parameters you (or the AI agent) supplied.
- Read-only. Every tool is a strictly read-only data lookup. None of them can write to Amazon, place orders, post reviews, modify listings, or take any side-effecting action on third-party platforms.
- HTTPS-only transport. Both the stdio variant (local) and the hosted variant (
https://mcp.pangolinfo.com/mcp) require HTTPS; HTTP requests are refused. - Open source. The full source is in this repository under MIT license — anyone can audit what the server sends and where.
- Responsible use. Pangolinfo APIs aggregate public e-commerce data. You are responsible for using the returned data in compliance with the terms of service of the underlying platforms (Amazon, Google, etc.) and with applicable laws in your jurisdiction.
Report security issues privately to [email protected] — please do not file public GitHub issues for vulnerabilities.
Support
- API docs: https://docs.pangolinfo.com
- Issues: https://github.com/pangolinfo/pangolinfo-mcp/issues
- Email: [email protected]
License
MIT © Pangolinfo
For synchronous product details (POST /api/v1/scrape, parserName=amzProductDetail), an explicit confirmation that the ASIN is missing on the requested marketplace returns 9204 with the message ASIN not found on this marketplace: B08HSS37H7 (the requested ASIN). Normal format-specific pricing applies; see pointCost (usually 1 credit for JSON). Do not automatically retry the same marketplace and ASIN; check the marketplace and identifier. The ASIN may exist on another marketplace. Network failures or no content remain 9200 without a charge. This describes the synchronous product-detail endpoint. get_amazon_product and scrape_url surface this as [ASIN_UNAVAILABLE], retriable=no.
Reviews (0)
Sign in to leave a review.
Leave a reviewNo results found