Skip to content

Latest commit

 

History

History

1_Steam_Reviews

readme.md

Steam Game Reviews Analyzer

Overview

A Python script that analyzes Steam game reviews using RAG (Retrieval Augmented Generation) and an agent workflow to provide comprehensive review summaries.

Prerequisites

  • OpenAI API key (store in .env file)
  • Python with required dependencies

Usage

  1. Open steam_reviews.py
  2. Set your target game's Steam URL in the STEAMURL variable:
STEAMURL = "https://store.steampowered.com/app/294100/RimWorld/"
  1. Run the script to generate a review summary

How It Works

The script follows a structured workflow:

  1. Retrieves the 100 most recent Steam reviews
  2. Filters for high-quality, informative reviews
  3. Stores filtered reviews in a vector database
  4. Collects game metadata (description, tags, genres)
  5. Generates a comprehensive summary using an agent with specialized tools

Known Limitations

The Steam Reviews API may occasionally:

  • Return empty JSON responses even for games with reviews
  • Not function properly with demo pages

Example of API limitation:

API endpoint: https://store.steampowered.com/appreviews/2867690?json=1&filter=recent&num_per_page=100

Response: {"success":1,"query_summary":{"num_reviews":0,"review_score":0,"review_score_desc":"Brak recenzji użytkowników","total_positive":0,"total_negative":0,"total_reviews":0},"reviews":[],"cursor":"*"}