steam-games-mcp

steam-games-mcp

Enables interaction with Steam: search games, get store details, reviews, prices, discounts, news, and player profiles, libraries, and achievements via the Steam Web API.

Category
Visit Server

README

steam-games-mcp

npm version CI license: MIT

An MCP server for Steam: search games and read store details, prices, reviews, discounts and news (no key), plus player profiles, libraries and achievements via the official Steam Web API (free key).

Works with any MCP client (Claude Desktop/Code, Cursor, VS Code, Cline, …) over stdio.

Once it's connected, just ask your agent in natural language.

No credentials needed (store, search & discovery):

"Find Hollow Knight and tell me its price, genres and age rating."
"What are recent reviews saying about Baldur's Gate 3?"
"Have Cyberpunk 2077's reviews recovered since launch?"
"Which games are >80% off right now with 90%+ positive reviews?"
"Which recent, well-reviewed games run on Steam Deck?"
"What's discounted on Steam's front page right now?"
"Show me Steam's top sellers and newest releases."
"Any recent patch notes for No Man's Sky?"
"How rare is each achievement in Elden Ring?"
"How many people are playing Counter-Strike 2 right now?"
"Get current prices for appids 620, 413150 and 1145360."
"For appids 1245620 and 1086940, show price, review % and Steam Deck status."

With a free API key + your STEAM_ID (your account; see Getting your credentials):

"Show my Steam profile."
"List my games by playtime."
"What have I played in the last two weeks?"
"What's on my wishlist that's discounted and well-reviewed?"
"List Hollow Knight's achievements and how rare each one is."
"How far am I through Elden Ring's achievements?"
"What's the SteamID64 for the profile name 'gabelogannewell'?"

Install

Add it to your MCP client's config. Store/search tools work with no credentials; player tools need a free Steam Web API key.

{
  "mcpServers": {
    "steam": {
      "command": "npx",
      "args": ["-y", "steam-games-mcp"],
      "env": {
        "STEAM_API_KEY": "your-steam-web-api-key (optional — enables player tools)",
        "STEAM_ID": "your-steamid64-or-vanity-name (optional — default 'you' for player tools)",
        "STEAM_COUNTRY": "US (optional — store price region)",
        "STEAM_LANGUAGE": "english (optional — store language)"
      }
    }
  }
}

Replace each value with your own; remove the optional lines you don't need. A free key comes from https://steamcommunity.com/dev/apikey. From source: npm ci && npm run build, then use "command": "node", "args": ["/ABS/PATH/steam-games-mcp/dist/index.js"]. As a .mcpb bundle: grab it from the latest release.

Getting your credentials

Store, search and discovery tools need nothing — skip this section if that's all you want. The player tools (profile, library, achievements, your wishlist) need a free API key and a public profile. Three short steps:

  1. Get a free Steam Web API key. Sign in at https://steamcommunity.com/dev/apikey, enter any domain (e.g. localhost), and copy the key into STEAM_API_KEY.

  2. Find your Steam ID. Set STEAM_ID to either:

    • your vanity name — the custom part of your profile URL steamcommunity.com/id/YOUR_NAME → YOUR_NAME (resolved automatically), or
    • your 17-digit SteamID64 (steamcommunity.com/profiles/7656…; or look it up at https://steamid.io).

    With STEAM_ID set you can just ask "my wishlist / library" — without it, give the agent a SteamID64 each time (use resolve_vanity_url to convert a name).

  3. Make your profile public (for your own library/achievements): Steam → profile → Edit Profile → Privacy Settings → set My profile and Game details to Public.

The key and Steam ID go in your MCP client config (the env block above) — see docs/clients.md for per-client examples. Never commit them.

What it does

Key: – no credentials · K Steam Web API key.

Tool Key Purpose
search_games – Find games by title → appid (with price)
get_game – Store details by appid or name: price, genres, platforms, Metacritic, age rating, DLC, requirements
get_items – Batch store card (price, review %, Steam Deck, release) for a LIST of appids — one call
discover_games – Find games catalog-wide by discount, recency, Steam Deck and rating — no appids needed
get_game_reviews – Review summary + recent reviews
get_review_histogram – Review trend over time (history + recent)
get_prices – Batch current price/discount for many appids
get_specials – Steam front-page discounts
get_featured – Featured sections (top sellers, new releases, …)
get_game_news – Recent news / patch notes
get_global_achievements – Global achievement unlock rates (rarity)
get_current_players – Live concurrent player count
get_wishlist – A player's wishlist appids (public profiles)
get_game_achievements K Full achievement list (names, descriptions) + rarity
resolve_vanity_url K Custom profile name → SteamID64
get_player_summary K Player public profile
get_owned_games K A player's games + playtime
get_recently_played K Games played in the last two weeks
get_player_achievements K A player's achievement progress in a game

Two tiers. Store/search + discovery tools (store/api.steampowered.com) need no credentials — including catalog-wide discovery (discover_games: deals, new releases, Steam Deck, rating) and batch price/review checks (get_items). Player tools need a free STEAM_API_KEY and a public profile; they return a clear message when the key is unset. Set STEAM_ID (a SteamID64 or vanity name) to make those tools default to you, so "my wishlist / library" works without passing an ID each time.

No third-party services: deal discovery and reviews come from Steam's own (keyless) store APIs. SteamDB is not used (no public API + scraping disallowed). Steam has no price-history API, so that isn't offered. Not affiliated with Valve.

Develop

npm install
npm run build        # type-check + bundle to dist/index.js
npm test             # node:test (mocked, offline)
npm run lint
npm run format
npm run check:api    # live upstream health-check (Storefront keyless; player check needs STEAM_API_KEY)
npm run inspector    # run under the MCP Inspector

Runtime requires Node ≥ 18. Contributor/agent guidance: AGENTS.md. Per-client config and all tunables: docs/clients.md.

Updating

  • npx: unpinned npx -y steam-games-mcp fetches the latest on the next run.
  • .mcpb bundle: download the new bundle from the releases page and reinstall.
  • From source: git pull && npm ci && npm run build.

License

MIT © Grinv

Recommended Servers

playwright-mcp

playwright-mcp

A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.

Official
Featured
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.

Official
Featured
Local
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.

Official
Featured
Local
TypeScript
VeyraX MCP

VeyraX MCP

Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.

Official
Featured
Local
graphlit-mcp-server

graphlit-mcp-server

The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.

Official
Featured
TypeScript
Kagi MCP Server

Kagi MCP Server

An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.

Official
Featured
Python
Neon Database

Neon Database

MCP server for interacting with Neon Management API and databases

Official
Featured
E2B

E2B

Using MCP to run code via e2b.

Official
Featured
Exa Search

Exa Search

A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.

Official
Featured
Qdrant Server

Qdrant Server

This repository is an example of how to create a MCP server for Qdrant, a vector search engine.

Official
Featured