Steam (Read Only)
Pull read-only Steam player counts, reviews, playtime, and catalog signals into agent research for game ideas.
Overview
Steam MCP is a MCP server for the Idea phase that provides read-only Steam Web API data for games, players, reviews, and live counts inside your agent.
What is this MCP server?
- Read-only Steam Web API: friends, owned games, playtime, achievements
- Sales, reviews, and live concurrent player counts for market scouting
- PyPI steam-mcp v1.9.0 with uvx runtime hint and stdio transport
- Requires STEAM_API_KEY from steamcommunity.com/dev/apikey
- No write actions—safe for querying public and key-scoped Steam data
- PyPI steam-mcp version 1.9.0
- Transport: stdio; runtimeHint uvx
- STEAM_API_KEY required (read-only API scope)
Community signal: 1 GitHub stars.
What problem does it solve?
Validating a Steam game idea means manually collating player counts, reviews, and catalog stats across sites while your agent sits idle.
Who is it for?
Indie game developers and agent users doing Steam market and audience research before building.
Skip if: Builders with no Steam API key, non-game products, or workflows that need to modify Steam accounts or store listings.
What do I get? / Deliverables
After setup, you can ask your agent for Steam-backed research answers in one thread while scoping art, scope, and genre fit.
- Agent-queryable Steam profile, library, and social read APIs
- Live and historical-style game signals for research prompts
- Version 1.9.0 server package on PyPI
Recommended MCP Servers
Journey fit
How it compares
Read-only Steam data MCP, not a game engine plugin or a Steam publishing automation suite.
Common Questions / FAQ
Who is Steam MCP for?
Solo builders and small teams researching Steam games, genres, and player activity through an MCP-capable coding agent.
When should I use Steam MCP?
Use it during idea and validate research when you need playtime, reviews, achievements, or live player counts to inform scope.
How do I add Steam MCP to my agent?
Obtain a STEAM_API_KEY, install steam-mcp from PyPI (uvx supported), configure stdio in your MCP client, and restart the agent.