Google Maps has no public reviews API, so browser automation is the only route, and keeping the surface to "fetch N reviews for this place" is the right scope for a scraper. Expect the usual fragility that comes with it, and note that the Excel export is what makes the results useful beyond a single conversation.
A small scraping server for Google Maps reviews. You hand the assistant the Maps link for a place and say how many reviews to pull; it drives a browser through Playwright to collect them, and the reviews come back for summarising. Collected reviews can also be written to an Excel file.
Scraping a chosen number of reviews from a place's Google Maps page so an assistant can summarise them, and saving the collected reviews to an Excel file for use elsewhere. The author lists retrieval-augmented search and stored review sets as intended additions rather than current behaviour.
Nothing — no account, no key. `uv` for dependency management, with `uv sync` from a clone; the package is `google-maps-reviews-mcp` (0.2.0 in pyproject). Playwright needs its system dependencies installed with `playwright install-deps`. The documented client entry points at your virtualenv's Python and the `server.py` file, over stdio.