# Changelog of App Review Teardown & Insights (AI) (`scrapebench/reviews-insight-mcp`) Actor

- **URL**: https://apify.com/scrapebench/reviews-insight-mcp/changelog.md
- **Full Actor documentation**: https://apify.com/scrapebench/reviews-insight-mcp.md

## Changelog

All notable changes to this actor are documented here. This renders on the Apify Store
Info tab, so keep entries buyer-readable.

### \[0.1.2] - 2026-08-24

- **Fixed: runs that came back with no reviews for an app that plainly has them.** Apple throttles
  its review feed per IP, and a shared cloud container's own IP is often already spent — it answers
  with an empty feed rather than an error, so the run looked like an app nobody had reviewed. Runs
  now go through Apify Proxy by default, which resolves it. You can still turn the proxy off.

### \[0.1.1] - 2026-08-24

- **An app that could not be analyzed now says why.** When the App Store review feed is
  temporarily unavailable (it has a daily outage window), when no store matches your query, or
  when the AI step fails, the row carries a `notice` explaining it instead of arriving empty —
  which read exactly like "this app has no reviews". Those rows are free, as they always were.
  This includes the case where the feed answers normally but hands back no reviews at all for an
  app the App Store rates — that is the feed being unavailable, not an app nobody has reviewed.

### \[0.1.0] - Unreleased

- First release: **AI competitive teardown** from a mobile app's reviews. Scrapes the **Apple App
  Store** (official iTunes feed) and, best-effort, **Google Play**, then synthesizes a structured
  teardown — summary, strengths, weaknesses, top complaints, feature requests, themed sentiment.
  Give several apps to compare competitors side by side.
- **Change monitoring:** with `compareToPrevious`, each run diffs against the previous one and
  reports rating delta, **emerging vs resolved complaints**, and sentiment shift — schedule it to
  track competitor apps over time.
- Charged per teardown (`ai-insight`); apps with no reviews are not charged. Also callable as an
  MCP tool by AI agents.
- **Multi-app comparison + sales battlecards:** give 2+ apps and get an extra record with a
  strengths/weaknesses matrix and battlecards (their weaknesses, where you win, watch-outs). Name
  one app as `yourApp` to frame the battlecards from your side. Charged as one more `ai-insight`.