# Analyze App Reviews and Prepare Approved Replies

App reviews scatter complaints, feature requests, and praise across individual messages. This skill uses ASOScan insights and review evidence to prioritize actions and prepare individually approved replies.

## Install

```bash
npx skillstore add asoscan/review-insights
```

## Metadata

- Status: approved
- Slug: asoscan-review-insights
- Version: 1.3.7
- Author version: 1.3.7
- Skillstore revision: r1
- Version status: valid
- Tree hash: 283de1954a9d37e799595cec9e1d3e17f62fcd194374e07cc0f4fa142160e9c8
- Author: asoscan
- GitHub username: asoscan
- Repository: https://github.com/asoscan/aso-skills/tree/026a04de63174b90ec1c14432dd845e86c1d9d6e/skills/review-insights
- Ref: 0ea67940dbf32cfb16450cfc953be73d0004de39
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: allowed
- Manual install advisory: allowed
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Risk factors: external\_commands, network, env\_access
- Quality score: 75
- Public page: https://skillstore.pages.dev/skills/asoscan-review-insights
- Manifest: https://skillstore.pages.dev/api/skills/asoscan-review-insights/manifest

## Capabilities

- Summarize review sentiment with positive, neutral, and negative shares and an average score shown on its -1 to \+1 range.
- Rank recurring topics, feature requests, and reported bugs by count, then recommend responses for the top three themes.
- Retrieve paginated raw reviews, filter by star rating, and quote review bodies verbatim as evidence.
- Analyze apps in the ASOScan account and competitors already tracked by that account.
- Read saved reply templates and request AI reply drafts after disclosing costs and obtaining consent.
- Post one developer reply only after approval of its exact text, subject to the app having a store connection.

## Use Cases

- Prioritize Product Improvements: Review frequent complaints and requests, inspect supporting reviews, and identify three concrete roadmap or onboarding actions.
- Prepare Individual Review Replies: Start from saved templates, request a draft after cost approval, and approve the exact response before publication.
- Study Tracked Competitor Feedback: Inspect sentiment and recurring review themes for an already tracked competitor to identify messaging and product opportunities.

## Prompt Templates

### Summarize Review Sentiment

```
Summarize review sentiment for [app]. Show the score range and analyzed count, and disclose pending reviews.
```

### Investigate Frequent Complaints

```
Find the top three complaints for [app]. Retrieve supporting one-star reviews and quote them exactly. Recommend an owner and response for each theme.
```

### Prepare a Review Reply

```
Find review [review ID] for [app] and offer a saved template. Ask before paid drafting, then show the draft without posting.
```

### Compare Feedback and Prioritize Actions

```
Compare [my app] with [tracked competitor] using sentiment and theme counts. Quote supporting reviews and flag incomplete analysis. Recommend three actions without changing data.
```

## Limitations

- Live data requires connected ASOScan tools or shell access with an existing ASOSCAN\_API\_KEY; neither connection is included in this skill.
- Coverage is limited to account apps and tracked competitors; pending analysis can leave some reviews outside the insights.
- AI drafting costs 2 API credits and 1 AI credit; posting costs 2 API credits and requires exact-text approval.
- Posting requires a store connection in ASOScan, and developer replies are not a search-ranking signal.

## Best Practices

- Show analyzed and pending counts, and always include the -1 to \+1 sentiment score range.
- Support priorities with retrieved reviews and preserve quoted review text exactly.
- Keep credentials private, approve chargeable drafts first, and approve each reply text separately before publication.

## Anti Patterns

- Inventing reviews, presenting paraphrases as quotations, or treating pending reviews as already analyzed.
- Posting unapproved replies, applying one approval to multiple replies, or hiding drafting costs.
- Presenting developer replies as a search-ranking tactic or claiming access to untracked competitors.

## Security Audit

- Audited at: 2026-11-15T18:56:49.914\+00:00
- Summary: All 28 static findings are false positives involving Markdown formatting, intended ASOScan authentication, documentation links, or review retrieval. No evidence found of credential exfiltration, malicious shell execution, system reconnaissance, or prompt injection in the supplied file. The instructions require consent for AI operations and data changes, plus separate approval of each exact reply before publication.

## Stats

- Views: 0
- Downloads: 2
- Favorites: 0
- Popularity score: 0
