MCP Analytics

ai.mcpanalytics由 mcpanalytics.ai 发布官方发布者v1.0.8更新于 Oct 24, 2026

A statistical analyst in your AI chat: bring data and a question, get a citable, re-runnable report.

已验证Streamable HTTP可网页运行数据库数据与分析

概览

AI 生成的概览

远程 MCP 服务器，充当统计分析助手：上传数据、提出业务问题，获得可引用、可复跑的报告。

可以做什么
  • 上传、列出和搜索数据集，并可原地刷新已有数据集
  • 把问题路由到合适的路径（复用、重跑、快速回答、库工具运行、估算或定制构建），本身不执行也不计费
  • 执行数据快速读取、经过验证的库分析，或基于已确认的估算下单定制分析
  • 在运行前查看并调整报告估算、检查工具适配性、查找先例
  • 浏览已交付报告及其卡片和可分享链接，并跨所有已交付分析提问
  • 管理每周或每月的定期重跑、查看构建状态，以及查询组织数据仓库
适用场景
适合在对话中对自己的数据集做统计或业务分析，得到标明方法、含图表、可在新数据上复跑的报告。它面向数据问题，而非通用助手。
运行要求
远程 Streamable HTTP 端点，无需本地安装包。需要 MCP Analytics 账号，并通过 X-API-Key 请求头提供 API 密钥；无密钥时仅能以发现模式列出工具。仓库查询需在商业版套餐中启用。
安装前请注意
API 密钥是敏感凭据，通过 X-API-Key 请求头发送。上传的数据会发送给第三方服务；README 称数据仅用于回答你的问题，保留期限见隐私政策。部分路径会产生付费工作：定制构建需要用户明确确认和幂等键，应先审阅估算。定期计划会创建重复运行并通过邮件发送报告。

安装

在 SourceWeft 中

  1. 打开 控制台中的 MCP Analytics，将其添加到工作区。
  2. 连接 X-API-Key 凭据。它们会被私密存储，绝不会在公开页面上展示。
  3. 为需要使用其工具的对话启用该服务。

Web executable，通过 Streamable HTTP。 远程服务在工作区中配置后即可从网页运行时运行。

交给智能体安装

把这段提示词粘贴到 Claude Code、Codex、Cursor 等智能体，或 SourceWeft 对话中。智能体会读取该服务器的安装指南，先向你展示发布者、端点和登录方式，经你同意后再把它添加到你的客户端。密钥不会写进任何命令。

阅读 https://sourceweft.com/mcp/ai.mcpanalytics%2Fanalytics/install.md，并按其中说明把这个 MCP 服务器接入我的客户端。修改任何配置之前，先告诉我它的发布者、端点和登录方式，等我确认。不要把 API 密钥写进命令或我会提交的文件；问我密钥应从哪里获取。

查看智能体所依据的安装指南

其他 MCP 客户端

把它添加到你客户端的 mcpServers 配置中。

{
  "mcpServers": {
    "analytics": {
      "type": "http",
      "url": "https://api.mcpanalytics.ai/mcp/api-key",
      "headers": {
        "X-API-Key": "<YOUR_API_KEY>"
      }
    }
  }
}

README

MCP Analytics

A statistical analyst in your AI chat. Bring a dataset and a question, get back a citable, re-runnable report with its method named.

MCP Analytics is a Model Context Protocol server that answers business and statistical questions about your data, as deep as each question needs:

  • Answer: a quick question about the data, read from all of it, shown with how we got it, and saved.
  • Analytics: a tested statistical method runs on the data, and we interpret it with charts and a written report.
  • Custom Analytics: a question only your business asks, built for your data from proven models, yours to re-run on next month's file.

Intelligent routing picks the depth: you describe the question, and the server says which path answers it and why. Analytics is repeatable: same data, same numbers. Every report names its method and the code that ran it, and runs again on new data.

It is for business and statistical questions about data, not a general-purpose assistant.

Website: mcpanalytics.ai · Free tools you can try in the browser: mcpanalytics.ai/free

Connect

Endpoint: https://api.mcpanalytics.ai/mcp/api-key (Streamable HTTP, JSON-RPC)

Authentication: an API key in the X-API-Key header. Create an account at mcpanalytics.ai and copy your key from your account settings.

json
{  "mcpServers": {    "mcpanalytics": {      "url": "https://api.mcpanalytics.ai/mcp/api-key",      "headers": { "X-API-Key": "mcp_your_key_here" }    }  }}

Claude Code:

bash
claude mcp add --transport http mcpanalytics https://api.mcpanalytics.ai/mcp/api-key --header "X-API-Key: mcp_your_key_here"

Browse before you sign up. Without a key the server answers in discovery mode: you can connect and list the tools, but not run anything.

bash
curl -X POST https://api.mcpanalytics.ai/mcp/api-key \  -H 'Content-Type: application/json' \  -d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}'

Start

Ask your AI assistant a question about your data, for example:

  • "Upload sales.csv. What changed in revenue from May to June, and which channel drove it?"
  • "Which customers are most likely to churn?"
  • "Forecast next quarter's sales from this file."

The first call is the resolver: it reads your question and your data, executes nothing and charges nothing, and names the path that answers it (an answer, a library analysis, or a custom build) with the reason. Then it runs that path.

What you get

  • A report with charts, tables and a written interpretation, with its method named and the code that ran it.
  • Re-runs on new data: run the same analysis on next month's file.
  • A library of your reports you can search and ask questions across.

Data and privacy

Your uploaded data is used only to answer your questions. See the privacy policy for what we keep and for how long. Security issues: see SECURITY.md.

Support

License

See LICENSE.

来源：README.md，提交 33a3580

工具

27

about

Platform documentation and info: how it works (Answer, Analytics, Custom Analytics), usage.

输入参数topic必填

agent_advisor

AI help desk: which analysis fits your question, interpreting results, fixing errors. Multi-turn.

输入参数message必填

datasets_upload

Get your data in. Pass `data` as an array of row objects to create the dataset immediately and get a dataset_ref ready for path_resolver; omit it to get an upload link for a file only the user can reach. Add replace_ref (uuid://ID:KEY) with data to REFRESH an existing dataset in place; schedules and tools holding that reference read the new data on their next run.

输入参数dataexpires_inreplace_ref

datasets_list

List and search your uploaded datasets, with fuzzy matching on name, description, and tags. Returns each dataset's uuid:// reference for use in path_resolver and path_tool_run.

输入参数limitsearch

discover_tools

Browse the analyses you can run: the ones you commissioned plus the platform Standard Library (prebuilt tools; each result tagged source:'own' or 'standard_library'). Plain-language match; no query lists everything, your own first. Nothing fits? Commission it with path_resolver.

输入参数query

tools_schema

Get an analysis's parameter schema. ALWAYS call before path_tool_run.

输入参数tool_name必填

review_estimate

The estimate as you review it WITH the user: the question as understood, the estimated answer (sample, marked), every place and its question, the page link, a review checklist. Before path_bespoke.

输入参数estimate_id必填

adjust_estimate

Apply the user's layout wishes to the estimate's page through the layout agent; a new named arrangement, nothing overwritten, no number changes.

输入参数estimate_id必填instructions必填

find_precedent

Before estimating: how did we answer this objective before, on this data or any data? Prior packages and library runs with their tools, mappings, bespoke module names, method and verdicts. Platform-wide, read-only.

输入参数kobjective必填dataset_ref

check_tool_fit

Before naming a library tool: does it fit THIS dataset for THIS question? Column mapping, missing required inputs, method-fit verdict, the places it delivers. Read-only.

输入参数objectivetool_name必填dataset_ref必填

package_status

Read an analytics package back: status, every run under it, the report link once delivered.

输入参数package_id必填

my_objects

List and search the objects you own across every question: the curated charts, tables and figures of each delivered package, grouped by objective.

输入参数limitqueryinclude_dropped

build_status

Check a commissioned build in-chat: stage progress, queue position, rejection reason if the data didn't match the objective, honest ETA, report link when delivered.

输入参数pipeline_idtrack_token

ask_library

Ask a question across all your delivered analyses: a synthesized answer with citations back to specific reports.

输入参数question必填

reports_list

Your report library: every analysis delivered, with status and links. Pass semantic_query to search report content in plain language.

输入参数limitsemantic_query

warehouse

Query your org's data warehouse free: browse the catalog (tables with column roles + computed metrics), semantically find data, plain-language ask, or named templates. Requires warehouse enablement (business plans).

输入参数queryaction必填paramsquestion

schedules

Standing re-runs of analyses you own: action='create' (weekly/monthly against a re-runnable data reference, connector:// or an https:// link; report emailed after each run), 'list', or 'cancel'.

输入参数action必填cadencetool_namedataset_refschedule_idcolumn_mapping

reports_view

Get a shareable browser link for a report, viewable without authentication.

输入参数processing_id必填

report_cards

Browse a delivered report's individual cards (charts, tables, insights) inline in chat.

输入参数processing_id必填

path_resolver

Which path a question about data should take (reuse, rerun, answer, tool_run, estimate, bespoke, ask or none), with the reason, every path's score and time, and the ready input for the path it picks. Executes nothing and charges nothing.

输入参数doormetaobjective必填dataset_ref

path_reuse

Hand back the report already delivered for this question on this data. Time: as path_resolver quotes.

输入参数objective必填dataset_ref必填resolution_id

path_rerun

Run a delivered analysis again, on its own dataset or a new one: the same tools and mappings, no rebuild. Time: as path_resolver quotes.

输入参数package_id必填dataset_refresolution_id

path_answer

A read of the data (a total, an average, a count, the top group), computed on every row, in this response. When the question is not a read it says so and names the next step. Time: as path_resolver quotes.

输入参数objective必填dataset_ref必填resolution_id

path_tool_run

Run one or more tested library tools on this dataset, with charts and a written report. No custom build. Time: as path_resolver quotes.

输入参数objective必填tool_names必填dataset_ref必填estimate_idresolution_idcolumn_mappingidempotency_keymodule_parameters

path_estimate

A rough answer over a sample and the layout of the full report, for the person to confirm before anything is ordered. Time: as path_resolver quotes.

输入参数objective必填tool_namesdataset_ref必填completenessresolution_idlayout_objective

path_bespoke

Order a custom analysis built for this data from a confirmed estimate, optionally with library tools alongside. Needs the person's yes and an idempotency key; a retry returns the first order. Time: as path_resolver quotes.

输入参数tool_namesestimate_id必填resolution_ididempotency_key必填layout_objective

版本历史

2
  1. v1.0.8最新Oct 15, 2026
  2. v1.0.7Sep 16, 2026

评分

已满足适用检查项的 100%

该评分衡量此收录条目的完整程度，依据注册中心条目、端点与公开来源计算得出。它不是对服务器本身的评价，也不表示服务器安全或当前可用：可连通性是带日期的检查结果，评分也不包含安装量或用户评价。

  • 近 30 天内可连通必需25 分
  • 已声明认证方式必需10 分
  • 已列出工具必需15 分
  • 描述必需15 分
  • 图标10 分
  • 文档15 分
  • 许可证10 分