App Pain API

io.github.apppainapi由 GitHub 帳號 apppainapi 發布v0.3.1更新於 Nov 1, 2026

What users hate about ANY iOS/Android app: clustered complaints, evidence, trends. No key needed.

已驗證Streamable HTTP可網頁執行網頁搜尋與抓取資料與分析商業與電商

概覽

AI 產生的概覽

遠端 MCP 伺服器，把任意 iOS/Android 應用程式的公開評論整理成排序後的痛點、趨勢與版本評分。

可以做什麼
  • 在 App Store 與 Google Play 搜尋應用程式並取得其 id
  • 將近期負評聚類成附嚴重度與原文引用的痛點排行
  • 比對歷史分析，看出哪些抱怨是新增、上升、下降或已解決
  • 讀取應用程式中介資料、評分分布與各版本的評論統計
  • 分頁取得原始評論以便深入查看
  • 管理競品監控清單，把新的 1-2 星評論與上升中的痛點推送到 webhook
適用情境
適合競品研究、產品機會發掘與應用程式拆解：想了解某款應用程式使用者最不滿什麼，或抱怨隨版本如何變化，而不必人工逐條讀評論時使用。
執行需求
只需遠端 streamable HTTP 端點，不需本機執行環境或安裝套件。無憑證時可用免金鑰的每日額度；選用的 x-api-key 標頭（註冊後取得）提供個人配額。需要能連線到託管伺服器，中國大陸使用者可能需要代理。
安裝前請注意
選用的 x-api-key 標頭屬於密鑰，不應寫入共用設定。競品監控與更大配額需付費方案（Starter、Pro），watch_competitors 會把評論摘要傳送到你設定的 Slack、Discord 或任意 https webhook。冷分析需 30-120 秒；所有工具據稱皆為唯讀。

安裝

在 SourceWeft 中

  1. 開啟 儀表板中的 App Pain API，將其新增到工作區。
  2. 為需要使用其工具的對話啟用該服務。

Web executable，透過 Streamable HTTP。 遠端服務在工作區中設定後即可從網頁執行環境執行。

交給代理程式安裝

把這段提示詞貼上到 Claude Code、Codex、Cursor 等代理程式，或 SourceWeft 對話中。代理程式會讀取該伺服器的安裝指南，先向你展示發布者、端點和登入方式，經你同意後再把它新增到你的客戶端。金鑰不會寫進任何命令。

閱讀 https://sourceweft.com/mcp/io.github.apppainapi%2Fapp-pain-api/install.md，並按其中說明把這個 MCP 伺服器接入我的客戶端。修改任何設定之前，先告訴我它的發布者、端點和登入方式，等我確認。不要把 API 金鑰寫進命令或我會提交的檔案；問我金鑰應從哪裡取得。

檢視代理程式所依據的安裝指南

其他 MCP 客戶端

把它新增到你客戶端的 mcpServers 設定中。

{
  "mcpServers": {
    "app-pain-api": {
      "type": "http",
      "url": "https://hawkmill.com/mcp"
    }
  }
}

README

[App Pain API logo]

App Pain API

What users hate about any iOS/Android app — clustered complaints, evidence quotes, trends. One MCP server, no API key.

[MCP] [Streamable HTTP] [No key required] [SealKeeper Service SEAL]

Live server · Docs · OpenAPI 3.1 · Pain index (101 apps) · Pricing

Review intelligence for any app on the App Store or Google Play — not just your own. Feed it an app id, a store link, or a bare package name; get back ranked pain-point clusters mined from real 1–2★ reviews, each with its share of complaints, a severity score, and verbatim evidence.

Built for competitive research and agent workflows: one tool call replaces a week of reading reviews by hand.

json
{  "painPoints": [    { "rank": 1, "title": "Intrusive ads & ad frequency", "sharePercent": 16, "severity": 4 },    { "rank": 2, "title": "Shadowbanning & algorithm suppression", "sharePercent": 12, "severity": 4 }  ],  "overview": "Monetization pressure (ads) and opaque algorithm changes dominate complaints."}

↑ real output — TikTok, US App Store. Cached analyses return instantly; ~100 popular apps are pre-analyzed and always free to read.

The tools

ToolWhat it does
search_apps ★Find apps on both stores (merged results, rich fields) — the step 1 of every teardown
mine_app_pain_points ★The core tool: fetch recent reviews, keep the negative ones, cluster them into ranked pain points with evidence quotes
get_pain_trendCompare two snapshots: what got worse, what got fixed (new / resolved / rising / falling)
get_version_historyRating per release — catch the version that broke things
get_app_reviewsRaw reviews, unified schema across stores (dev replies, thumbsUp)
get_app_detailsMetadata: rating histogram, category, price, IAP
watch_competitorsPaid: push new 1–2★ reviews and rising pain points to Slack/Discord/webhooks

Also ships two prompts (competitor_pain_report, app_pain_deep_dive) so an agent can run a full teardown from one sentence.

Install

Server URL — paste it into any MCP client, no key needed:

https://app-pain-api.quiet-delta.workers.dev/mcp
ClientHow
Claude Codeclaude mcp add --transport http app-pain-api https://app-pain-api.quiet-delta.workers.dev/mcp
Claude Desktop / claude.aiSettings → Connectors → Add custom connector → paste the URL
ChatGPTSettings → Apps & Connectors → Developer mode → Create → paste the URL, auth: No authentication
Cursor~/.cursor/mcp.json → {"mcpServers":{"app-pain-api":{"url":"https://app-pain-api.quiet-delta.workers.dev/mcp"}}}
VS Codecode --add-mcp '{"name":"app-pain-api","type":"http","url":"https://app-pain-api.quiet-delta.workers.dev/mcp"}'
Windsurf~/.codeium/windsurf/mcp_config.json → {"mcpServers":{"app-pain-api":{"serverUrl":"…/mcp"}}}
Codex CLI~/.codex/config.toml → [mcp_servers.app-pain-api] url = "…/mcp"
Gemini CLI~/.gemini/settings.json → {"mcpServers":{"app-pain-api":{"httpUrl":"…/mcp"}}}
stdio-only clientsnpx -y mcp-remote https://app-pain-api.quiet-delta.workers.dev/mcp

Keyless limits: unlimited cached results plus a few new analyses per day. Want your own quota (10/month free, more on paid tiers)? Register at POST /v1/register and add an x-api-key header. When the server-side allowance runs out, mine_app_pain_points degrades gracefully — it hands your agent the prepared negative reviews to cluster with its own model.

Mainland China: *.workers.dev needs a system proxy or TUN mode to connect.

Try asking your agent

  • "What do users hate most about Duolingo, and is it getting better?"
  • "Compare the top 4 habit-tracker apps' pain points and find a product gap."
  • "Watch my 3 biggest competitors and alert our Slack when their users start complaining about something new."

Also a plain REST API

Everything the MCP server does is available over REST with JSON responses and a full OpenAPI 3.1 spec — same cache, same quota, one GET:

bash
curl -s "https://app-pain-api.quiet-delta.workers.dev/v1/apps/ios/835599320/painpoints"

Pricing

Free tier forever (no credit card) · Starter $9/mo (any country, 500-review depth, watch 2 competitors) · Pro $29/mo (daily alerts for 15 competitors, unlimited analyses) — details on the site.

Notes

  • Data comes from public app-store pages, normalized into one schema. Not affiliated with Apple or Google.
  • All tools are read-only (readOnlyHint annotations set).
  • Speaks the classic initialize handshake (2024-11-05 → 2025-11-25) and answers the mandatory server/discover RPC from the 2026-07-28 spec — modern clients probe, see the supported version list, and fall back to the handshake automatically.
  • Cold analyses take 30–120s and stream progress notifications; repeat calls hit a 7-day cache and return in milliseconds.
  • Privacy · Terms · Acceptable use
  • Contact: [email protected]

來源：README.md，提交 e83da69

工具

7

search_apps

Search apps

Search the iOS App Store and Google Play (both merged by default) to find apps and their ids. Use this FIRST when you only know an app name or a category.

輸入參數q必填limitstorecountry

mine_app_pain_points

Mine app pain points

THE core tool: fetch recent reviews of any iOS/Android app, keep the negative ones, and cluster them into ranked pain points (title, sharePercent, severity 1-5, verbatim evidence quotes). Ideal for competitive research and product discovery. Cached 7 days (instant on repeat); a COLD server-side analysis takes 30-120s. If the result has "mode":"agent", it contains prepared negative reviews and "instructions" — do the clustering yourself and present it to the user.

輸入參數appId必填modelstoreanalyzecountryreviewsmaxRating

get_pain_trend

Pain-point trend

How an app's pain points changed over time: compares two archived analyses and labels each theme new / resolved / rising / falling / stable with share deltas (percentage points). Instant (archive read, no LLM). Default: latest vs ~30 days earlier. Returns a note when the app has fewer than two snapshots yet.

輸入參數tofromappId必填store

get_app_details

App details

App metadata: name, developer, category, price, average rating, rating count and 1-5★ histogram (when available).

輸入參數appId必填storecountry

get_version_history

Version history

Per-version review stats (review count, average rating, first/last seen) — spot a rating crash right after a bad release. Android: aggregated from recent reviews. iOS: current version + release notes.

輸入參數appId必填storecountry

watch_competitors

Watch competitors (alerts)

Continuous competitor monitoring (paid plans: Starter 2 apps weekly, Pro 15 apps daily). Manage the watchlist and the alert destination; digests of NEW 1-2★ reviews and NEW/RISING pain points are pushed to a Slack / Discord / any https webhook. action: list | add | remove | set_webhook | test.

輸入參數appIdstoreaction必填webhookUrl

get_app_reviews

Raw reviews

Fetch raw app reviews (unified schema across stores: rating/title/body/version/date/devReply/thumbsUp), 50 per page. Use for drill-down after mine_app_pain_points.

輸入參數pagesortappId必填storecountry

版本歷史

3
  1. v0.3.1最新Nov 1, 2026
  2. v0.3.0Oct 25, 2026
  3. v0.2.0Oct 23, 2026

評分

已滿足適用檢查項的 90%

該評分衡量此收錄條目的完整程度，依據註冊中心條目、端點與公開來源計算得出。它不是對伺服器本身的評價，也不表示伺服器安全或目前可用：可連通性是帶日期的檢查結果，評分也不包含安裝量或使用者評價。

  • 近 30 天內可連通必需25 分
  • 已宣告認證方式必需10 分
  • 已列出工具必需15 分
  • 描述必需15 分
  • 圖示10 分
  • 文件15 分
  • 授權條款10 分

    請在儲存庫或註冊中心條目中宣告授權條款。