API reference · Reviews
Scan from a sentence
Say what reviews you want in a sentence ("bad reviews for gym apps", "crash complaints in budgeting apps since the last update"). A model reads it into a filter (the ratings, the words to look for, how recent) and both stores are searched with several phrases for the apps it means, the apps more phrases found ranked first. Nothing is pulled: the result carries `pull`, the body for POST /reviews/pulls, to send as it is or change first. `readBy` is `rules` when the model was unavailable and simple word rules read it; check the filter before pulling then. Under a minute.
/api/v1/reviews/scancurl "https://appleak.app/api/v1/reviews/scan" \
-H "Authorization: Bearer $APPLEAK_KEY" \
-H "Content-Type: application/json" \
-d '{
"query": "bad reviews for gym apps since the last update",
"country": "us"
}'{
"id": "6f1c2b9e-1d0a-4c55-9a51-2b7f0c3e8d41",
"object": "job",
"kind": "review_scan",
"status": "queued",
"credits": 1,
"refunded": false,
"progress": null,
"createdAt": "2026-10-19T12:00:00.000Z",
"startedAt": null,
"finishedAt": null,
"error": null,
"result": null,
"links": {
"self": "/api/v1/jobs/6f1c2b9e-1d0a-4c55-9a51-2b7f0c3e8d41"
}
}#Parameters
Body parameters
querystringrequiredWhat reviews you want, in plain words. 2–600 characters
countrystringoptionaldefault"us"Two-letter storefront code, e.g. us, gb, de.
forcebooleanoptionalStart a new run even when the same one is already queued or running for you.
#Returns
A job, straight away, with status queued. Follow it with GET /jobs/{id} until it is done; the finished job carries the result.
MCP: start_review_scan — Start an Appleak review scan: turns a sentence like "bad reviews for gym apps" into a review filter (ratings, keywords, recency) and finds the App Store and Google Play apps it means. Pulls nothing. Spends credits (one question); returns a job id. get_job gives the filter, the apps, and `pull`: the input for start_review_pull.