原型 · 预览版:收录官方注册中心里的全部服务器,每 30 分钟同步一次。评价和账户功能将在后续阶段推出。
PopNx MCP为使用 AI 的人准备的 MCP 服务器 · 用大白话讲清楚

ANIME INTELLIGENCE / AI PURCHASE INTELLIGENCE

作者:Mike05102版本 3.10.2i★ –i在云端运行
注册名称io.github.Mike05102/anime-intelligence-mcp

安装

概览

作者提供的说明(英文原文)

AI commerce intelligence for anime, industrial parts, replacements, sourcing and agent buying.

PopNx MCP 的三大支柱

这三项永远不会合并成一个分数:意见、测量和人气是不同的东西。

它能帮你做什么

我们还没有为它写通俗的介绍。下面是我们从注册中心读取到的事实。

  • 连接方式: 在云端运行
  • 传输方式: Streamable HTTP
  • 分类: 游戏、体育、美食与学习

安装

选择你使用的 AI 应用,然后复制设置。

安装之前

你需要

你用的是哪个应用?

在上方选择你的应用。我们会记住它,下次按钮会显示“安装到 …”。

适用于网页版、桌面版和手机版 Claude:

  1. 打开 Settings → Connectors
  2. 点击 Add custom connector
  3. 粘贴下面的地址,然后点击 Add
  4. 如果要求登录,请登录作者的服务
SettingsConnectors1+ Add2https://…3
https://anime-intelligence.goodmy0312.workers.dev/mcp

在终端中运行这条命令

claude mcp add --transport http anime-intelligence https://anime-intelligence.goodmy0312.workers.dev/mcp

在 Cursor 中打开

或者自己复制设置

粘贴到项目中的 .cursor/mcp.json,或 ~/.cursor/mcp.json

{
  "mcpServers": {
    "anime-intelligence": {
      "url": "https://anime-intelligence.goodmy0312.workers.dev/mcp"
    }
  }
}

浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。

在 VS Code 中打开

或者自己复制设置

粘贴到项目中的 .vscode/mcp.json

{
  "servers": {
    "anime-intelligence": {
      "type": "http",
      "url": "https://anime-intelligence.goodmy0312.workers.dev/mcp"
    }
  }
}

浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。

ChatGPT 只能连接互联网上的服务器:

  1. 打开网页版 ChatGPT → Settings → Connectors (Apps)
  2. 点击 Create 或 Add(部分套餐需要先在 Advanced 中开启 Developer mode)
  3. 粘贴下面的地址,起个名字,然后保存
  4. 如果要求登录,请登录作者的服务
SettingsConnectors1+ Create2https://…3
https://anime-intelligence.goodmy0312.workers.dev/mcp

菜单名称可能因套餐和版本不同而略有差异。

你的选择通过 Cookie 保存在这台设备上,无需账户。

这些设置是根据作者在注册中心的条目生成的。如果 README 的写法不同,请以 README 为准。

它能做什么

我们还没有读取这个服务器的工具列表;我们会轮流检查各个端点。

基础检查i

我们检查的是任何人都能看到的信息;我们没有阅读代码。不适用的检查项不计入总数。 · 检查于 2026年11月8日

  • ✓ 通过

    作者与仓库一致

    注册名称已通过 GitHub 账户 Mike05102 验证,该账户拥有此仓库

  • … 尚未检查

    仍在维护

    我们还没有从 GitHub 读取这个仓库

  • … 尚未检查

    有许可证

    我们还没有从 GitHub 读取这个仓库

  • … 尚未检查

    端点有响应

    尚未检查;我们会轮流检查各个端点

  • … 尚未检查

    使用 OAuth 登录

    尚未检查;我们会轮流检查各个端点

  • … 尚未检查

    工具描述中没有隐藏指令

    尚未检查;我们会在检查端点时读取工具列表

与“游戏、体育、美食与学习”中的其他服务器比较i

  • 星标数多于此分类中其他 1292 个服务器的 0%
  • 发布到注册中心的时间比此分类中其他 1292 个服务器的 47% 更新
  • 基础检查结果等于或优于此分类中其他 1292 个服务器的 100%

与注册中心里这个分类中所有活跃的服务器比较。只有当发布者拥有该仓库时,星标数才计入。

常见问题

ANIME INTELLIGENCE / AI PURCHASE INTELLIGENCE 安全吗?

它在 1 项适用的基础检查中通过了 1 项。我们没有阅读代码。在添加密钥或令牌之前,请先看看这个服务器要求哪些权限,并使用只具备所需权限的密钥。

它需要哪些密钥或代码?

注册中心没有列出必需的密钥,不过有些服务器会在首次使用时要求你登录。

它在我的电脑上运行还是在云端?

它在作者的云端运行。通过互联网连接,无需安装,但你发送的内容会经过作者的服务器。

它还在维护吗?

GitHub 仓库最后更新于–,最新版本(3.10.2)于32天前进入注册中心。

哪些应用可以使用它?

支持 MCP 的应用,例如 Claude Desktop、Claude Code、Cursor 和 VS Code。它通过 URL 连接,所以接受远程 MCP 服务器的应用也可以使用,例如开启开发者模式的 ChatGPT。

有什么可以替代它?

同一分类中有:像塑 MCP Plus, Pixelbolt pixel art toolkit, monty by Nuvi

替代选择

同一分类中其他作者的服务器,按星标数排序。

如何看卡片:↓ 每周下载量i★ GitHub 星标i三大支柱:评价 · 检查 · 使用情况i更新 = 最新版本发布到注册中心的时间i

像塑 MCP Plus

Unofficial macOS MCP for Douyin AR (像塑): assets, menus, AI combine, delivery. For learning only.

作者:Muqian-Sun非常新今天更新

在你的电脑上运行

Pixelbolt pixel art toolkit

Pixel art and sprite animation for agents: JSON recipes to PNG, APNG, GIF and sprite sheets.

作者:skulitom非常新今天更新

在你的电脑上运行

monty by Nuvi

AI video editing for Uzbek creators: reels, lessons, podcast clips, captions.

作者:monty.uz起步阶段今天更新

在云端运行✓ monty.uz

作者的 README

打开 README

按原文从作者的仓库复制。PopNx MCP 没有审查或背书其内容。HTML、图片和脚本已被移除。链接会离开本站。

ANIME INTELLIGENCE / AI PURCHASE INTELLIGENCE

Version 3.10.2

Production MCP endpoint:

https://anime-intelligence.goodmy0312.workers.dev/mcp

Production API base:

https://anime-intelligence.goodmy0312.workers.dev

Repository:

io.github.Mike05102/anime-intelligence-mcp


Overview

ANIME INTELLIGENCE / AI PURCHASE INTELLIGENCE is an MCP + API + x402 commerce-intelligence service designed for autonomous AI agents.

The service currently has three main capability groups:

  1. Anime / collectible purchase intelligence
  2. PARTS INTELLIGENCE for industrial, replacement and discontinued parts
  3. Generic commerce search for larger product markets

Travel adapters also exist in the codebase, but live travel inventory remains disabled until provider credentials are configured.

The service is built on:

  • Cloudflare Workers
  • Supabase
  • MCP
  • OpenAPI
  • x402
  • Solana USDC
  • Yahoo Shopping integration
  • eBay Browse integration
  • Rakuten affiliate routing
  • public merchant / purchase pages

The design goal is to let an AI agent move from:

search → identify → evaluate → decide → select seller → pay for intelligence → hand off to purchase

without requiring a human to manually assemble information from multiple sources.


1. Anime / Collectible Intelligence

The original ANIME INTELLIGENCE engine focuses on physical Japanese anime collectibles.

Supported product categories include:

  • scale figures
  • Nendoroid
  • figma
  • prize figures
  • plush
  • acrylic goods
  • keychains
  • badges
  • model kits
  • trading cards
  • collaboration apparel
  • collaboration sneakers
  • other physical anime merchandise

The system is designed to handle difficult identity problems such as:

  • multiple editions of the same character
  • rereleases
  • manufacturer differences
  • Japanese-only product names
  • JAN/EAN differences
  • prize vs retail versions
  • seller-title noise
  • incomplete marketplace listings
  • ambiguous multilingual search queries

The service resolves a canonical product before returning higher-value purchase intelligence.


Free Anime Search

GET /v1/search

Example:

/v1/search?query=Hatsune%20Miku

Use this for candidate discovery when the exact product is not yet known.


2. Paid Anime x402 Intelligence

Paid endpoints use Solana USDC via x402.

Current paid endpoints:

EndpointPurposePrice
/v1/identifyexact product / edition identification0.005 USDC
/v1/marketcurrent market value / price context0.01 USDC
/v1/rarityscarcity and rerelease risk0.01 USDC
/v1/authenticitycounterfeit / listing mismatch risk0.02 USDC
/v1/buy-waitBUY / WAIT / WATCH / AVOID decision0.02 USDC
/v1/best-placebest current purchase route0.03 USDC
/v1/listing-matchlisting-to-canonical product verification0.01 USDC
/v1/deadlinepreorder / lottery / purchase deadline0.01 USDC
/v1/landed-costbuyer-country-aware landed cost0.02 USDC
/v1/price-historyhistorical price context0.02 USDC
/v1/full-intelligencecomplete collectible decision intelligence0.05 USDC
/v1/shopping-intelligenceagent-ready purchase recommendation0.05 USDC

The service resolves identity before payment when possible so an agent does not pay for the wrong product.


x402 Payment Flow

  1. Agent calls a paid endpoint without payment proof.
  2. Server returns HTTP 402.
  3. Agent reads the PAYMENT-REQUIRED response header.
  4. Agent creates the required Solana USDC payment.
  5. Agent retries the exact same URL with PAYMENT-SIGNATURE.
  6. X-PAYMENT is also accepted as a compatibility alias.
  7. After verification and settlement, the endpoint returns HTTP 200.
  8. Successful responses include PAYMENT-RESPONSE.

Current x402 network:

  • Solana mainnet
  • USDC
  • x402 v2

3. PARTS INTELLIGENCE

Version 3.10.x adds a second specialized information business:

PARTS INTELLIGENCE

It targets industrial, replacement, maintenance, discontinued and difficult-to-source parts.

Example categories:

  • PLC
  • sensors
  • relays
  • switches
  • servo motors
  • servo amplifiers
  • inverters
  • power supplies
  • pneumatic components
  • cylinders
  • valves
  • bearings
  • motors
  • connectors
  • industrial control components
  • machine replacement parts
  • appliance / service parts
  • camera parts
  • audio parts
  • tool parts
  • IT / server replacement parts

The commercial value comes from resolving questions such as:

  • What exact part is this?
  • Is it discontinued?
  • What is the successor?
  • Is there a verified replacement?
  • Is a candidate truly compatible?
  • Which official document supports the relationship?
  • Is the original still available?
  • What is the best sourcing route?

PARTS Data Model

PARTS INTELLIGENCE uses Supabase as the system of record.

Core tables:

parts_catalog

Canonical part identity.

Typical fields:

  • manufacturer
  • brand
  • part number
  • normalized part number
  • name
  • category
  • lifecycle status
  • discontinued date
  • successor part number
  • specifications
  • dimensions
  • electrical attributes
  • mechanical attributes
  • official URL
  • source confidence
parts_relations

Explicit relationships between parts.

Supported relation types:

  • successor
  • substitute
  • compatible
  • incompatible
  • used_in

This table is one of the most important assets in the service.

parts_documents

Evidence and official reference documents.

Examples:

  • official PDF
  • catalog
  • manual
  • datasheet
  • specification sheet
  • replacement table
  • compatibility note
parts_offers

Observed sourcing information.

Examples:

  • merchant
  • price
  • currency
  • condition
  • stock status
  • URL
  • observation time
parts_sources

Configured official manufacturer sources.

The initial source registry contains:

  • OMRON
  • Mitsubishi Electric
  • SMC
  • IDEC
  • Panasonic Industry
  • Yaskawa
  • Oriental Motor
  • THK
  • NSK
  • Azbil

PARTS Compatibility Policy

PARTS INTELLIGENCE does not infer compatibility merely because:

  • product names look similar
  • dimensions look similar
  • model numbers look similar
  • marketplace sellers claim interchangeability

A positive substitute / compatibility recommendation requires an explicit relation record with supporting evidence.

Unverified candidates are returned separately.

This is a deliberate safety and data-quality rule.


PARTS Endpoints

GET /v1/parts/status

Returns PARTS database and ingestion status.

GET /v1/parts/search

Free part search.

Example:

/v1/parts/search?part_number=E2E-X5ME1

or:

/v1/parts/search?query=OMRON%20sensor
GET /v1/parts-intelligence

Paid full PARTS intelligence.

Price:

0.05 USDC

Typical output includes:

  • canonical part identity
  • lifecycle state
  • discontinuation status
  • successor relationship
  • verified substitute relationships
  • unverified candidate relationships
  • incompatibility evidence
  • official documents
  • sourcing observations
  • purchase / sourcing decision

4. PARTS Ingestion Mode in v3.10.2

Version 3.10.1 included a larger autonomous manufacturer-site crawler.

Version 3.10.2 intentionally removes that crawler from the Cloudflare Worker runtime to reduce deployment size and runtime complexity.

Current ingestion mode:

Supabase bulk import + admin upsert

This keeps the production Worker smaller and more stable.

Initial target:

  • several thousand high-quality parts
  • then approximately 10,000 high-value FA / industrial parts
  • then expand by manufacturer and category

No new external vendor API key is required for the PARTS core.

Existing Yahoo / eBay integrations may optionally enrich sourcing information.


PARTS Database Setup

Run:

PARTS_INTELLIGENCE_v3_10_2_SCHEMA.sql

once in the existing Supabase SQL Editor.

The schema creates:

  • parts_catalog
  • parts_relations
  • parts_documents
  • parts_offers
  • parts_sources
  • parts_ingest_runs

It also seeds the initial official manufacturer source registry.


5. Generic Commerce Search

The server also exposes a more general AI purchase-intelligence layer.

POST /v1/commerce/search

Supported domains currently include:

  • consumer_electronics
  • automotive_parts
  • enterprise_procurement
  • mro_and_industrial_components
  • electronic_components

Current provider integrations reuse existing configured commerce sources.


Identity Guards

Different domains use different recommendation rules.

Consumer electronics

Generic market search is allowed.

Automotive parts

An exact OE / MPN / part number is required before recommendation.

Enterprise procurement / MRO / electronic components

An exact manufacturer part number is required before recommendation.

The service deliberately avoids recommending technically incompatible components from broad keyword similarity alone.


6. Travel Adapters

Travel code remains present in the Worker.

Current architecture supports:

  • Booking.com Demand
  • Duffel
  • Expedia Rapid
  • Amadeus adapter slot

Current production status:

awaiting_credentials

No travel provider is considered live unless:

GET /v1/travel/status

reports the provider as configured.

Travel is not currently the focus of the service because those providers require external API credentials.


7. MCP

MCP endpoint:

https://anime-intelligence.goodmy0312.workers.dev/mcp

The server supports Streamable HTTP MCP.

Relevant tools include:

  • search_anime_product
  • paid anime intelligence tools
  • record_purchase_intent
  • commerce_search
  • parts_search
  • parts_intelligence
  • parts_status
  • travel tools

MCP metadata is designed so AI agents can understand:

  • when to use a tool
  • why the tool is paid
  • expected output
  • x402 payment behavior
  • identity requirements
  • compatibility restrictions

8. OpenAPI and Discovery

OpenAPI:

/openapi.json

LLM-readable service description:

/llms.txt

x402 discovery metadata:

/.well-known/x402

MCP discovery metadata:

/.well-known/mcp.json

Agent profile:

/agent/profile

Agent services:

/agent/services

Health:

/health

9. Public Purchase Pages

Rakuten affiliate URLs are not directly redistributed through API responses.

Instead, the API can return a public product page hosted by ANIME INTELLIGENCE.

Example pattern:

/shop/product/<canonical-product-id>

The public page can contain merchant routing and affiliate links in a human-readable media context.

This separation reduces ambiguity between machine-facing intelligence and merchant-facing affiliate routing.


10. Telemetry and KPI

The service records conversion telemetry in Supabase.

Important x402 funnel events include:

  • API request
  • canonical product selection
  • x402 gate entered
  • payment challenge issued
  • payment retry received
  • payment attempt
  • payment verified
  • paid call
  • settlement success
  • settlement failure

Important commerce events include:

  • purchase intent created
  • merchant selected
  • purchase URL served
  • affiliate click
  • checkout started
  • purchase confirmed

External Revenue KPI

The system separates:

  • true external commercial traffic
  • crawlers / monitors
  • admin tests
  • known self-tests
  • contextless synthetic probes

Important target KPIs include:

confirmed_external_shopping_intent_calls > 0

external_canonical_product_selected > 0 payment_retry_received > 0 external_payment_attempts > 0 external_payment_verified > 0 external_paid_calls > 0 external_revenue_usdc > 0

A self-payment from the operator's own wallet should not be treated as external commercial revenue.


11. Health Check

Production health endpoint:

https://anime-intelligence.goodmy0312.workers.dev/health

A healthy deployment should currently report:

{

"ok":

只显示了开头部分;其余内容请在仓库中阅读

问 PopNx