mcpField
S
Steam Review Mcp

Steam Review Mcp

Local
github.com/fenxerVersion Control7Updated 2d ago
Repository

About Steam Review Mcp

Steam Review MCP lets LLMs retrieve Steam game reviews and store descriptions. Its single review tool returns paginated review data alongside game details, and includes prompts for summarizing overall or recent player feedback.

Key features

  1. 01Retrieve a page of game reviews
  2. 02Return review scores and totals
  3. 03Clean review text and provide a next-page cursor
  4. 04Fetch store name and description
  5. 05Provide prompts for review analysis

Use cases

  1. 01Summarize a game's review strengths and drawbacks
  2. 02Examine recent player feedback
  3. 03Retrieve review pages for a Steam app ID

MCP support

The MCP server exposes the `get_steam_review` tool, which fetches reviews and store information for a Steam app ID. It also provides prompts for overall and recent review analysis.

#steam-review#mcp#game-reviews

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

mcp.json · Universal config
{
  "mcpServers": {
    "steam-review": {
      "command": "npx",
      "args": [
        "-y",
        "steam-review-mcp"
      ]
    }
  }
}
Run commandnpx -y steam-review-mcp

Local servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.

Packages
npmsteam-review-mcp

Tools

1

Saved tool list · updated 3w ago

ParameterTypeRequiredDescription
appidstringYesSteam application ID
filterstringNorecent: sorted by creation time, updated: sorted by last updated time,all: (default) sorted by helpfulness, with sliding windows based on day_range parameter, will always find results to return.
languageenumNoLanguage filter (e.g. english, french, schinese). Default is all languages.all | arabic | bulgarian | schinese | tchinese | czech | danish | dutch | english | finnish | french | german | greek | hungarian | indonesian | italian | japanese | koreana | norwegian | polish | portuguese | brazilian | romanian | russian | spanish | latam | swedish | thai | turkish | ukrainian | vietnamese
day_rangenumberNorange from now to n days ago to look for helpful reviews. Only applicable for the all filter.
cursorstringNoreviews are returned in batches of 20, so pass * for the first set, then the value of cursor that was returned in the response for the next set, etc. Note that cursor values may contain characters that need to be URLEncoded for use in the querystring.
review_typestringNoall:all reviews (default), positive: only positive reviews, negative: only negative reviews
purchase_typestringNoall: all reviews, non_steam_purchase: reviews written by users who did not pay for the product on Steam,steam: reviews written by users who paid for the product on Steam (default)
num_per_pagenumberNonumber of reviews to get, max 100, default 50

Details

From the project README

Steam Review MCP

English | 中文

Access Steam game reviews using Model Context Protocol (MCP).

MCP Badge

Install MCP Server

README image

Features

Helps LLMs retrieve Steam game reviews and information:

  • Get one page of game reviews (scores, totals, cleaned review text, and the next cursor)
  • Get the store name and full description from IStoreBrowseService/GetItems
  • Analyze game reviews and summarize pros and cons

Installation

Installing via Smithery

To install Steam Review for Claude Desktop automatically via Smithery:

npx -y @smithery/cli install @fenxer/steam-review-mcp --client claude

Run it directly with npx:

npx steam-review-mcp

or add：

{
  "mcpServers": {
    "steam-review-mcp": {
      "command": "npx",
      "args": [
        "steam-review-mcp"
      ]
    }
  }
}

Usage

Tools

This MCP service provides the get_steam_review tool. Pass a Steam app id and the tool reads:

store.steampowered.com/appreviews was disabled on 2026-10-22. Parameters use the old string names (all, recent, steam) and the server maps them into input_json.

The tool returns text JSON and the same object as structuredContent:

  1. game_reviews:

    • success: true when Steam returned a response object
    • review_score: Review score, or null when that page has no score
    • review_score_desc: Review score description
    • total_positive: Total positive reviews
    • total_negative: Total negative reviews
    • total_reviews: Positive plus negative totals
    • total_matching: Reviews matching the filter
    • day_range_used: Day range the Helpful filter actually searched
    • cursor: Pass this back to fetch the next page
    • reviews: Cleaned review text for this page

  2. game_info:

    • name: Store name
    • detailed_description: Store description with BBCode removed

An empty Steam response means no matching reviews. Aggregates are null and reviews is empty. Set STEAM_WEB_API_KEY only when you need the higher rate limit. The public methods work without it.

Code mode

The server stays compatible with Cloudflare's single-tool code mode as an upstream SDK v1 server. get_steam_review publishes an output schema, read-only annotations, and structuredContent. A host can wrap this server with codeMcpServer(). This package does not replace its tools with a code tool, because that wrapper is host-side and current code mode still serves the legacy SDK v1 handler.

Prompts
summarize-reviews

For overall game review analysis, summarizing the pros and cons of the game.

Parameters
  • appid (required): Steam game ID, e.g., 570 (Dota 2)
recent-reviews-analysis

For analyzing recent game reviews, summarizing the current state of the game and player feedback.

Parameters
  • appid (required): Steam game ID, e.g., 570 (Dota 2)

Development

# Install dependencies
npm install

# Build project
npm run build

# Run service
npm start

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.