Google Reviews

Setup

On first use, read setup.md to prioritize company-level review research and then define optional monitoring cadence.

When to Use

Use this skill when the user wants to research review signals about any company across Google surfaces and decide quickly with evidence.

Use it for one-off company checks, competitor review comparisons, reputation due diligence, and source-based review analysis on Google Maps and Google Shopping.

If the user needs recurring tracking after the first analysis, switch into monitoring mode with heartbeat refreshes, sentiment trends, and scheduled reports.

Architecture

Memory lives in ~/google-reviews/. See memory-template.md for structure and status fields.

~/google-reviews/
|-- memory.md                     # Stable monitoring preferences and activation behavior
|-- brands/
|   `-- {brand}.md               # Per-brand scope, sources, and thresholds
|-- snapshots/
|   `-- {brand}/{source}.jsonl   # Normalized review snapshots by refresh cycle
|-- reports/
|   |-- daily/
|   `-- weekly/
`-- heartbeat/
    `-- monitor-state.md         # Last run timestamp, alert cooldowns, and health notes

Requirements

User provides:

• Company or brand targets and analysis question scope
• Access method for each source (official API, export, or user-approved fetch workflow)
• Optional alert channel and reporting cadence for recurring tracking

Optional tooling:

• jq for JSON shaping in shell workflows
• Spreadsheet or BI destination for long-range trends

Quick Reference

Data Storage

All skill-local monitoring state stays in ~/google-reviews/. Create on first use:

mkdir -p ~/google-reviews/{brands,snapshots,reports/daily,reports/weekly,heartbeat}

Core Rules

1. Start in Research Mode Before Monitoring

• Begin with the user question about a company: what they need to know, where, and why.
• Pull current review evidence first (ratings, review volume, theme mix, recency) for the requested company scope.
• Return a clear answer with sources and confidence before proposing any recurring workflow.

2. Normalize Every Source into One Review Schema

• Ingest source data through the canonical fields in review-schema.md.
• Keep source-native IDs and timestamps for traceability.
• Never merge records without dedup keys (source, entity_id, review_id).