知你AI助手｜多平台客户与客服数据 MCP

io.github.zhinikefuPublished by GitHub account zhinikefuv1.0.1Updated Oct 24, 2026

汇聚个微、企微、视频号、微信小程序、公众号、服务号、微信客服、微信小店、抖音号、小红书、微博、网站及H5客服的客户资料、会话与聊天记录，供AI查询分析、会话研判与受确认的客户触达。

VerifiedStreamable HTTPWeb executableCommunication & CollaborationData & AnalyticsBusiness & Commerce

Overview

AI-generated overview

Remote MCP server that lets an AI assistant query multi-channel customer profiles, sessions and chat history from Zhini customer service, and send confirmed…

What you can do
  • Look up the current agent identity, agent list, platform channels, tag groups and tags
  • Search customers by name, phone, WeChat ID, channel, owner, tag or time, and read profiles and tags
  • List active session queues and search historical sessions by business conditions
  • Read message context for a customer or session, including keyword and time-based lookups
  • Check API Key quota and usage
  • Open selected conversations in bulk, upload images, create, track, list and delete bulk message-sending tasks
When to use it
For support, sales and customer-operations teams already using Zhini customer service who want an assistant to triage waiting conversations, recall prior context, find refund or complaint threads, spot buying intent, or send a confirmed batch follow-up.
Requirements
A remote Streamable HTTP endpoint; no local package. Authentication is either MCP-native OAuth 2.1 browser authorization or a static Bearer API Key sent in the Authorization header (for example via the ZHINI_API_KEY environment variable), created after registering on the vendor site. Data access is limited to the permissions of the credential.
Before you install
The credential grants access to customer profiles and chat records, so store the API Key in environment variables or secure config and revoke OAuth grants when no longer needed. Bulk opening conversations and message sending are write actions that should only run after explicit user confirmation; sending is asynchronous, deletion of a task is irreversible, and already-sent messages cannot be recalled. Customer data is sent to the vendor's remote service.

Installation

In SourceWeft

  1. Open 知你AI助手｜多平台客户与客服数据 MCP in the dashboard and add it to a workspace.
  2. Connect the required credentials. They are stored privately and never shown on public pages.
  3. Enable the server for the chats that should use its tools.

Web executable via Streamable HTTP. Remote servers run from the web runtime once configured in a workspace.

Use with an agent

Paste this prompt into Claude Code, Codex, Cursor or another agent — or into SourceWeft chat. The agent reads this server's install guide, shows you its publisher, endpoint and sign-in, and adds it to your client once you agree. Keys never go into a command.

Read https://sourceweft.com/mcp/io.github.zhinikefu%2Fzhini-ai-assistant/install.md and connect the MCP server it describes to my client. Before changing any config, show me its publisher, endpoint and sign-in method and wait for my OK. Never put an API key in a command or in a file I would commit; ask me where it should come from.

Read the install guide the agent follows

Other MCP clients

Add this to your client's mcpServers config. The client completes OAuth sign-in on first connect.

{
  "mcpServers": {
    "ai": {
      "type": "http",
      "url": "https://mcp.h5bqb.top/mcp"
    }
  }
}

README

知你AI助手 MCP｜让 AI 查询多平台客户资料、研判会话并触达客户

将知你客服中的多平台客户资料、当前与历史会话及聊天记录连接到 Codex、Claude Code、WorkBuddy 等 AI 工具，通过自然语言完成客户查询、会话研判和业务分析；支持用户明确确认后批量打开选定对话、发送文本或图片消息。

知你AI助手官网 · 知你客服官网 · MCP 服务地址：https://mcp.h5bqb.top/mcp

产品介绍

知你AI助手是一项面向 AI 工具的多平台客户沟通数据连接服务。它通过 Streamable HTTP MCP，将已接入知你客服的客户资料、标签、负责人、当前会话、历史会话和聊天记录连接到 AI 工具。

团队可以直接用自然语言查找客户、了解沟通前情、检索退款或投诉记录、筛选目标客户、发现购买意向，并基于完整消息上下文整理业务结论；需要触达客户时，AI 在用户明确确认后批量打开选定对话、异步发送文本或图片消息并跟踪发送进度。查询范围由当前凭据权限决定，结果可以继续追溯到相关客户、会话和原始消息。

服务支持两种认证方式：MCP 原生 OAuth 2.1（浏览器授权，无需 API Key，含动态客户端注册与 PKCE，授权可随时在官网吊销）或 静态 Bearer API Key。

支持的平台渠道

个微/企微/视频号/微信小程序/公众号/服务号/微信客服/微信小店/抖音号/小红书/微博/网站及H5客服等平台渠道。

核心能力

  1. 多平台客户数据统一查询：查询已接入知你客服的个微、企微、视频号、微信小程序、公众号、服务号、微信客服、微信小店、抖音号、小红书、微博、网站及 H5 客服等平台渠道数据。
  2. 当前客服会话分诊：查看个微、企微、微信客服、微信小程序、公众号、服务号、网站及 H5 客服等渠道的待接待、我的会话、同事会话和 AI 会话。
  3. 客户资料与沟通前情查询：按姓名、手机号、微信号、渠道、客服、负责人、标签或时间，查找来自个微、企微、视频号、抖音号、小红书、微博等渠道的客户。
  4. 客户画像信息整理：查询微信小店、微信小程序、公众号、服务号及其他已接入渠道客户的资料、标签、来源渠道、负责人和联系信息。
  5. 历史会话与聊天记录检索：检索微信客服、个微、企微、抖音号、小红书、微博、网站及 H5 客服等渠道的历史会话。
  6. 消息关键词与时间查询：按退款、投诉、报价、试用、购买等消息关键词和发生时间，定位视频号、微信小店、抖音号、小红书等渠道的相关客户会话。
  7. 完整消息上下文读取：将 AI 整理结果追溯到对应平台渠道、客户、会话和原始消息。
  8. 渠道与业务字典查询：查询平台账号、渠道、客服和标签字典。
  9. API Key 用量查询：查询当前 API Key 的总额度、已使用量和剩余额度。
  10. 客户触达与消息发送：在用户明确确认后，批量打开选定客户对话（每次最多 20 个），创建最多 100 个 UID、最多 2 条文本/图片消息的异步群发任务，并跟踪发送进度、查询和管理历史任务。

MCP 工具

工具用途
zhini_get_current_kefu查询当前调用者对应的客服身份
zhini_list_active_sessions查询当前客服会话队列
zhini_get_customer_profile查询客户资料与标签
zhini_list_tag_groups查询标签分组
zhini_list_tags分页查询标签
zhini_search_tags按关键词搜索标签
zhini_fetch_messages读取客户或会话的消息上下文
zhini_list_kefu查询客服列表
zhini_list_channels查询平台渠道列表
zhini_search_customers按业务条件搜索客户
zhini_search_sessions按业务条件搜索历史会话
zhini_get_apikey_usage查询 API Key 用量
zhini_open_sessions将用户明确选定的对话批量打开到"我的对话"（≤20/次）
zhini_upload_image_for_message_sending上传消息发送图片并返回 CDN 地址（≤10 MiB）
zhini_create_message_sending_task两阶段创建批量消息发送任务（≤100 UID、≤2 条消息）
zhini_get_message_sending_task_status查询发送任务进度与逐客户明细
zhini_list_message_sending_tasks分页查询历史群发任务
zhini_delete_message_sending_task逻辑删除群发任务并取消未发送明细

使用方法

1. 连接 MCP（两种认证方式二选一）

方式一：OAuth 2.1 浏览器授权（推荐）——不配置任何静态认证字段，客户端首次访问时按服务端 OAuth 挑战打开浏览器完成知你账号授权：

json
{  "mcpServers": {    "zhini": {      "type": "http",      "url": "https://mcp.h5bqb.top/mcp"    }  }}

方式二：静态 API Key——访问知你AI助手官网，注册或登录后创建 API Key：

json
{  "mcpServers": {    "zhini": {      "type": "http",      "url": "https://mcp.h5bqb.top/mcp",      "headers": {        "Authorization": "Bearer ${ZHINI_API_KEY}"      }    }  }}

环境变量：

text
ZHINI_API_KEY=YOUR_API_KEY

2. 用自然语言查询与触达

  • 现在有哪些客户正在等待接待？
  • 查一下这个客户最近聊了什么，还有哪些事情待确认？
  • 找出近 7 天提到退款的历史会话。
  • 哪些客户最近提过报价、试用或购买？
  • 把这些已选客户的对话打开到我的对话。
  • 向已确认的客户批量发送跟进文本或图片，发完告诉我进度。
  • 整理最近登录异常反馈的共同问题。

适用团队

  • 客服团队：会话分诊、回复前情查询、投诉与问题复盘、批量跟进触达
  • 销售团队：客户背景查询、意向识别、跟进线索整理、确认后批量发送跟进消息
  • 客户运营：标签筛选、客户分群、跨渠道客户洞察
  • 质检与产品：历史问题检索、共性问题归纳、需求线索发现

连接信息

项目信息
产品知你AI助手 / Zhini AI Assistant
官网https://ai.zhinikefu.com/?f=github
MCP 地址https://mcp.h5bqb.top/mcp
传输方式Streamable HTTP
鉴权方式OAuth 2.1 浏览器授权 或 Bearer API Key
API Key 环境变量ZHINI_API_KEY
MCP 标识com.zhinikefu/zhini-mcp-service
npm 包@zhinikefu/zhini-mcp-service

数据与权限说明

  • 查询能力保持只读；批量打开对话与消息发送为受控触达操作，仅在用户明确确认后执行，且不修改客户资料、不修改标签、不接待/转接/关闭会话。
  • 消息群发是异步任务组合：先上传图片（如有），再创建发送任务，并通过状态、列表和删除接口管理任务；单任务最多 100 个 UID、最多 2 条消息，删除任务不可逆，已发送的消息不会撤回。
  • 查询结果受当前凭据对应的数据权限约束；OAuth 授权可随时在官网吊销。
  • API Key 应保存在客户端环境变量或安全配置中。
  • “微信”“抖音”“小红书”“微博”等名称用于说明已接入的数据来源渠道，不代表相关平台对本产品的官方背书。

关键词

多平台客户数据 MCP、客户资料查询 MCP、客服会话查询 MCP、历史会话分析、客户画像、客户分群、投诉会话分析、销售线索识别、消息群发、客户触达、批量发送、多平台客服、omnichannel customer data、customer conversation search、customer profile lookup、customer segmentation、message sending。

Source: README.md at commit cc28262

Tools

0
This server asks you to sign in, so SourceWeft could not read its tools without an account and has not indexed them. What a connected account sees can depend on that account and its permissions.

Version history

2
  1. v1.0.1LatestOct 14, 2026
  2. v1.0.0Sep 16, 2026

Score

75% of the applicable checks met

How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.

  • Reached in the last 30 daysRequired25 pts
  • Auth declaredRequired10 pts
  • Tools listed15 pts

    Answer tools/list without a credential, or ship a package that starts in a sandbox without one.

  • DescriptionRequired15 pts
  • Logo10 pts
  • Documentation15 pts
  • License10 pts

    Declare a license in the repository or the registry entry.