demandex-mcp
MCP server for Demandex e-commerce demand intelligence, providing Reddit-based opportunity cards and demand verdicts with free and paid per-call tools settled via x402/USDC.
README
demandex-mcp
MCP server for Demandex — pay-per-call e-commerce demand intelligence for AI agents. Demandex mines Reddit complaint/intent posts across ~70 communities into scored opportunity cards with verbatim evidence and permalinks, and answers ad-hoc demand verdicts for any physical-product query. Pay per call with USDC on Base via x402 — no signup, no accounts, no API keys.
- 2 free tools — no wallet, no API key.
- 5 paid tools — flat per-call USD prices, settled in USDC on Base Mainnet via x402.
Try before you pay: the free get_sample_opportunity tool returns a real, fixed opportunity
card in the same full shape as the paid get_opportunity, so you can learn the exact response
shape first.
Quick start
Free tier (no wallet)
{
"mcpServers": {
"demandex": {
"command": "npx",
"args": ["-y", "demandex-mcp"]
}
}
}
You get get_categories and get_sample_opportunity.
Paid tier (with wallet)
Add an EVM wallet private key (0x-prefixed) that holds USDC on Base Mainnet:
{
"mcpServers": {
"demandex": {
"command": "npx",
"args": ["-y", "demandex-mcp"],
"env": { "EVM_PRIVATE_KEY": "0x..." }
}
}
}
The paid tools then settle each call automatically.
Install snippets
Claude Code:
claude mcp add demandex -- npx -y demandex-mcp
# with a wallet:
claude mcp add demandex --env EVM_PRIVATE_KEY=0x... -- npx -y demandex-mcp
Claude Desktop / Cursor: add the JSON block above to your MCP config
(claude_desktop_config.json or .cursor/mcp.json).
Tools
Free tools
| Tool | Endpoint | Description |
|---|---|---|
get_categories |
GET /v1/categories |
Category index with opportunityCount, topScore, lastUpdated per category. |
get_sample_opportunity |
GET /v1/sample/opportunity |
Real, fixed sample opportunity card showing the full paid card shape. |
Paid tools
| Tool | Endpoint | Price (USD) | Description |
|---|---|---|---|
get_trending_opportunities |
GET /v1/opportunities/trending |
$0.01 | Top 20 trending cards as teasers (id, slug, title, category, score, signalCount, lastSeen). |
get_opportunities |
GET /v1/opportunities |
$0.02 | Up to 50 cards in a category (teaser + summary). Optional minScore (0-100). |
get_opportunity |
GET /v1/opportunity |
$0.05 | One FULL card by id or slug (exactly one) — all score components, productAngle, summary, verbatim evidence[]. |
gauge_demand |
POST /v1/gauge |
$0.03 | Demand verdict for a product query (3–200 chars), cached 7 days, corpus-only. Optional category. |
gauge_demand_live |
POST /v1/gauge/live |
$0.10 | Demand verdict with a LIVE Reddit search at request time (~45s budget). Same body as gauge_demand. |
Prices are flat per call — no bundles, no entitlements. The verdict shape (both gauges):
{ query, verdictScore (0-100), demandTemperature (cold|warm|hot), rationale, topEvidence[], relatedOpportunities[], cache, fetchedAt }.
How payments work (x402)
Demandex speaks the x402 exact scheme. The API's 402 advertises a
Base (USDC, EIP-3009) rail; this MCP settles on Base:
- The MCP requests a paid endpoint. With no payment, the API replies HTTP 402 with the accepted terms.
- If
EVM_PRIVATE_KEYis set, the MCP settles the call via the standard@x402/fetchV2 client and retries. If no key is set, the MCP returns the 402 price terms so you can see the cost without paying. - The API returns the data plus a
charged: truemarker.
Compute-first, settle-after: Demandex computes the full answer before charging, so you
are never billed for an error or an empty result (a still-warming index returns a no-charge
503 corpus_warming).
No accounts, no API keys — just a wallet with USDC on Base Mainnet. Physical products only.
Environment variables
| Variable | Required | Description |
|---|---|---|
EVM_PRIVATE_KEY |
No | 0x-prefixed wallet key with USDC on Base Mainnet. Required to pay the 5 paid tools on a live API. PRIVATE_KEY is accepted as a fallback. Free tools work without it; without a key the paid tools surface the 402 price terms. |
DEMANDEX_API_URL |
No | Override the API base URL (default https://api.demandex.dev). |
Disclaimer
Demandex provides informational e-commerce demand intelligence aggregated from public Reddit posts; physical products only; not professional or business advice; provided as-is. See https://api.demandex.dev/terms.txt.
Docs for agents: https://api.demandex.dev/llms.txt · full: https://api.demandex.dev/llms-full.txt
License
MIT © jcislo
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。