About Steam MCP
Steam MCP is an MCP server for game developers and community managers who need Steam market data in an AI coding assistant. It combines current store, review, and news data with historical player trends and a curated market digest.
Key features
- 01Search Steam apps
- 02View app, price, and release details
- 03Filter and analyze reviews
- 04Summarize news and patch notes
- 05Track player-count trends
- 06Get a curated market digest
Use cases
- 01Check a game's overall status
- 02Compare review sentiment and themes
- 03Review player trends for launch planning
- 04Follow market and competitor signals
MCP support
Exposes Steam search, app details, reviews, news, player trends, and market pulse through MCP tools. Supports stdio and Streamable HTTP connections.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
uvx playtide-steam-mcpSTEAM_MCP_BACKEND_URLOverride the companion backend base URL. Defaults to the Playtide hosted beta (https://api.playtide.dev).STEAM_MCP_AUTO_UPDATESet to 0 to disable the boot-time update check against the vendor update channel.STEAM_MCP_UPDATE_ORIGINPin a different update channel URL.Local servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.
Tools
Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.
Details
From the project READMEsteam-mcp
Steam market intelligence for AI agents — an MCP server that gives your coding assistant live Steam data: app search, player-count trends, store details, review mining, and news digests.
Built for game developers and community managers who want Steam answers inside their IDE conversation instead of tab-switching to store pages.
Why
Steam's public APIs are fragmented (store search, appdetails, user stats) and none of them keep player-count history — the single most useful signal for launch planning. steam-mcp stitches the surfaces together behind one tool set and pairs them with a companion backend that snapshots player counts every 30 minutes, so your agent can answer "how is the game actually doing" with real numbers.
Install
# from PyPI
pip install playtide-steam-mcp
steam-mcp
or from source:
git clone https://github.com/Playtide-studios/steam-mcp.git
cd steam-mcp
npm install
npm start
Requires Node 26+. Zero runtime dependencies — the server talks plain HTTP/JSON-RPC over stdio.
Tools
| Tool | What it does |
|---|---|
steam_search_apps | Resolve a game name to appids via Steam store search |
steam_get_app_overview | One-call health check: players, reviews, price, release, genres, latest news |
steam_get_store_details | Full store listing: pricing per region, platforms, DLC count, metacritic |
steam_get_app_reviews | Review mining with recency/sentiment/keyword filters + representative quotes |
steam_get_game_news | News and patch-note digests, summary or full text |
steam_get_player_trend | Player-count history with min/max/avg, week-over-week change, and a plain-language insight line |
steam_get_market_pulse | Curated market digest for game developers: Steam trends, live-ops patterns, competitor notes |
Client configuration
stdio (Claude Desktop / Cursor / any MCP client):
{
"mcpServers": {
"steam": {
"command": "node",
"args": ["/path/to/steam-mcp/src/index.ts"]
}
}
}
Streamable HTTP:
npm run start:http # serves MCP on http://localhost:8788/mcp
Ready-to-edit client configs live in examples/.
Configuration
| Variable | Default | Purpose |
|---|---|---|
STEAM_MCP_BACKEND_URL | https://api.playtide.dev | Companion backend base URL — hosted beta by default; override to self-host. Must be https:// unless loopback |
STEAM_MCP_BACKEND_KEY | — | Backend API key (optional; the hosted beta is open) |
STEAM_MCP_REMOTE_MANIFEST | 0 | Opt in to fetching tool descriptions from the backend each boot |
STEAM_MCP_TIMEOUT_MS | 10000 | Per-request timeout to Steam or the backend |
Everything else is zero-config: tool schemas are pinned to the packaged manifest (see src/fallback-manifest.json).
Data provenance & safety
- Store search, details, reviews, and news come directly from Steam's public endpoints at request time — nothing is scraped ahead or cached on disk.
- Trends and the market digest come from the companion backend.
- All free-text fetched from the internet (reviews, news bodies) is treated as data, never instructions — tool descriptions tell the model the same.
- The server sends no telemetry anywhere. The only outbound calls are the Steam endpoints you asked about and the backend you configured.
The companion backend
Trend and pulse data come from the steam-mcp companion backend — a small Express + Postgres service that snapshots player counts for every tracked app every 30 minutes.
-
Hosted backend (open beta — the default): installs use
https://api.playtide.devout of the box; no key needed during beta. To pin it explicitly:STEAM_MCP_BACKEND_URL=https://api.playtide.dev npm start -
Self-hosted: the backend is a standard docker-compose Postgres + API stack; set
STEAM_MCP_BACKEND_URLto your instance. See docs/TROUBLESHOOTING.md for connectivity help.
Troubleshooting
Backend unreachable, HTTPS requirement errors, deprecated_client on trend calls, rate limits — see docs/TROUBLESHOOTING.md.
Contributing
Issues and PRs are welcome — see CONTRIBUTING.md. Bug reports that include the failing tool call JSON get fixed first.
License
MIT © Playtide Studios. Steam is a trademark of Valve Corporation; this project is not affiliated with Valve.
Comments
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.