App review data, built for agents

App stores are hostile to scrapers, undocumented, and partitioned by country. ParseMyApp does that part for you and hands back structured JSON. Five tools, two prompts, one endpoint, no key.

MCP endpoint (Streamable HTTP)

https://www.parsemyapp.com/mcp

No auth

No API key, no account, no signup

Depth

Up to 500 reviews per app

Coverage

6M+ apps, 40+ countries, iOS + Android

Repeat calls

Cached — same app costs one scrape

Tools

search_app(term, store, country, limit)No model call

Resolve an app name to its App Store ID or Play Store package name.

get_reviews(app_id, store, country, limit)No model call

Raw reviews as JSON: author, rating, title, content, date.

get_quotes(app_id, store, country, theme, limit)No model call

Only the reviews mentioning a keyword. Deterministic substring match — typically 20x fewer tokens than fetching everything.

trend_diff(app_id, store, country, weeks)No model call

Week-over-week average rating, review volume, and direction. Pure computation over fetched reviews.

analyze_app(app_id, store, country, app_name, limit)Runs server-side, rate limited

The full structured analysis in one call: pain points, praises, opportunities, competitor gaps, contradictions — each with a verbatim quote, root cause, and frequency.

Prompts

analyze_app_reviews

Reason over raw reviews in your own model instead of ours. Same output schema, your framing.

compare_apps

Two apps side by side: shared pain points, where each has the edge, and whitespace neither covers.

Two ways to spend tokens

analyze_app runs the analysis on our model and returns finished JSON — one call, no reasoning tokens on your side, works unattended in a loop.

get_reviews plus the analyze_app_reviews prompt hands you the raw evidence to reason over yourself. More tokens, but you control the framing and can follow threads a fixed schema misses. Neither path needs an API key.

Agent skill

A drop-in skill file with the workflows worth knowing — competitor teardown, rating-drop post-mortem, pre-launch demand validation, ASO copy mining — plus the analysis rules and the limits of review data.

parsemyapp-skill.md

Setting this up in Claude or ChatGPT?

Connector setup guide