appstore-review-cli - App Store & Google Play Review Scraper

You have access to a CLI tool called appstore-reviews that scrapes Apple App Store and Google Play reviews. Use it when the user asks about competitor analysis, app reviews, user complaints, feature gaps, or bug reports from app stores.

You ARE the analyzer. Fetch reviews with the CLI commands below, then analyze them directly in your response. Do NOT suggest the user install Ollama or run the analyze command - you have the intelligence to do the analysis yourself.

Prerequisites

The tool must be installed before use. Run this to install and verify:

pip install appstore-review-cli && appstore-reviews --help

For Google Play support: pip install "appstore-review-cli[google]"

If appstore-reviews is still not found after install (PATH issue), use the fallback:

python -m appinsight search "app name"
python -m appinsight reviews <APP_ID> --stars 2

python -m appinsight works identically to appstore-reviews and does not require PATH setup.

Always use the CLI in the terminal. Do NOT use the Python API or import statements - use the bash commands shown below.

Important Behavior

• Output streams: Data goes to stdout, progress/status messages go to stderr. Use 2>/dev/null to suppress status messages when piping.
• Review limit: Apple's RSS feed returns a maximum of ~500 reviews per country (10 pages × 50 reviews). Don't expect more.
• Deduplication: Reviews are automatically deduplicated across pages.
• Input validation: --stars accepts 1-5, --pages accepts 1-10. Invalid values produce an error.
• Filters stack: --stars, --min-stars, --keywords, --days, and --version combine with AND logic. Each filter narrows the previous result.
• Rating range: --stars 2 means 1-2 stars. --min-stars 3 --stars 4 means 3-4 stars. --min-stars 3 --stars 3 means only 3-star reviews.
• Sorting: Default is newest first. Use --sort votes to surface the most impactful reviews, --sort rating for lowest-rated first.
• "No reviews match": If filters return zero results, the output will say "No reviews match the given filters." This is normal - try relaxing filters (fewer keywords, more days, higher star rating).
• No reviews at all: If an app has zero reviews in the store, the CLI will say so clearly and suggest trying a different country. Small or new apps often have no reviews.
• Network errors: If Apple's API is unreachable, the CLI prints a friendly error to stderr instead of a traceback.
• Google Play: Use --store google with package names (e.g. com.Slack). Requires pip install appstore-review-cli[google].
• Google Play search: The first search result sometimes lacks a package name (library limitation). Use the package name directly if needed (find it in the Google Play URL).

Output Rules

NEVER redirect CLI output to files (no > file.txt, > reviews.csv, etc.). Always read the output directly from stdout. The CLI is designed to print results to the terminal for you to analyze inline. Do not create .txt, .csv, .json, or any other files with review data unless the user explicitly asks for a file export.

Available Commands

1. Search for an app

appstore-reviews search "app name"

Returns app IDs, names, ratings. Use this first to find the numeric app ID.

Options:

• --limit 10 - Max results to return (default: 5)
• --format json - Output as JSON instead of table (useful for structured processing)

--country also works with search (e.g., --country germany to search the German App Store).

2. Fetch and filter reviews

appstore-reviews reviews <APP_ID> [options]

Options (all optional, filters stack with AND logic):

• --stars 2 - Max star rating to include (1-2 stars). Valid range: 1-5.
• --min-stars 3 - Min star rating to include (3+ stars). Use with --stars for exact range (e.g. --min-stars 3 --stars 3 for only 3-star reviews).
• --days 30 - Only reviews from the last 30 days
• --keywords crash,freeze,bug - Only reviews mentioning these words (case-insensitive, matches title or content)
• --version 5.0.1 - Only reviews for a specific version
• --pages 5 - Fetch more pages, 1-10 (default: 3, max useful: 10 = ~500 reviews)
• --format json|text|csv|markdown - Output format (default: text). Use csv for pandas/spreadsheet analysis.
• --sort date|rating|votes - Sort order: date (newest first, default), rating (lowest first), votes (most helpful first)
• --stats - Show rating distribution (printed to stderr, won't pollute data output)
• --country us - App Store region. Accepts codes (us, gb, de) or names (germany, japan, united kingdom)

Analysis Modes

When the user asks for analysis, fetch the relevant reviews with the CLI, then analyze them yourself in your response. Do not run analyze or pipe to another LLM - you are the LLM.

Compare Multiple Apps

Use the compare command for side-by-side analysis:

appstore-reviews compare <APP_ID_1> <APP_ID_2> --stars 2 --pages 5

Produces: overview table, per-app rating distribution, top complaint categories, top keywords, shared vs unique complaints.

Version Diff

Compare sentiment between app versions:

appstore-reviews version-diff <APP_ID> --pages 5
appstore-reviews version-diff <APP_ID> --old 4.23.0 --new 4.29.149

Shows: version comparison table, category changes with arrows, new/resolved issues, top keywords per version. Auto-detects the two most reviewed versions if --old/--new are omitted.

Trend

Show rating trend over time:

appstore-reviews trend <APP_ID> --pages 5
appstore-reviews trend <APP_ID> --period month --stars 2

Shows: per-period average rating, review count, trend arrows (▲/▼), sparkline bars, star distributions, and overall trend direction.

All three commands support --format json and --format csv for machine-readable export.

Gap Finder

Fetch 1-2 star reviews and analyze for unmet needs:

appstore-reviews reviews <APP_ID> --stars 2 --days 90 --pages 5 --format text

Look for patterns: "I wish it had...", "competitors do X better", "missing feature", "switched to Y".
In your response, group complaints by feature category, rank by frequency, and cite specific reviews as evidence.

Bug Hunter

Fetch 1-2 star reviews filtered to technical issues:

appstore-reviews reviews <APP_ID> --stars 2 --keywords crash,bug,freeze,error,broken,slow,stuck --format text

In your response, group by symptom (crashes, performance, data loss), identify affected versions, and rank by severity.

Sentiment Snapshot

Get the big picture before diving in:

appstore-reviews reviews <APP_ID> --stats --pages 5

Check the rating distribution first, then drill into the problem areas with targeted filters.

Example Workflows

"What are users complaining about in Slack?"

appstore-reviews search "Slack"
# Find the app ID from results (e.g., 803453959)
appstore-reviews reviews 803453959 --stars 2 --days 60 --format text --stats

Then categorize the complaints: UX issues, missing features, bugs, performance. Rank by how often each category appears.

"Find payment bugs in Spotify"

appstore-reviews reviews 324684580 --stars 2 --keywords payment,subscribe,charge,billing --format text

Group results by symptom: failed payments, unwanted charges, subscription cancellation issues.

"Compare negative reviews for two competing apps"

appstore-reviews compare <APP_A_ID> <APP_B_ID> --stars 2 --pages 5

The compare command produces a structured report with per-app breakdowns, shared complaints, and unique weaknesses. You can supplement with additional analysis.

"Check reviews for a specific country"

appstore-reviews reviews 803453959 --stars 2 --country germany --format text

"Find complaints about Spotify on Android"

appstore-reviews --store google reviews com.spotify.music --stars 2 --days 60 --format text

Group complaints by category: performance, features, bugs.

"How has Slack's rating been trending?"

appstore-reviews trend 803453959 --pages 5

The trend command shows weekly averages with sparklines. Use --period month for broader view.

Python API (for scripts/notebooks only)

Only use this if the user is explicitly writing Python code or a notebook - otherwise always use the CLI commands above.

from appinsight import get_reviews, get_reviews_df

# As list of dicts (no pandas needed)
reviews = get_reviews(618783545, stars=2, days=30)

# Google Play
reviews = get_reviews(\"com.Slack\", stars=2, days=30, store=\"google\")

# As pandas DataFrame (requires: pip install appstore-review-cli[pandas])
df = get_reviews_df(618783545, stars=2, pages=5)
df.groupby("version")["rating"].mean()
df[df["content"].str.contains("crash", case=False)]

Or export CSV from the CLI:

appstore-reviews reviews 618783545 --stars 2 --format csv