My-Actor: Pokemon Generator This Apify actor demonstrates a web scraping project built with Crawlee and Cheerio, enhanced with AI-powered Pokémon generation using the Google Gemini API. It scrapes data, generates Pokémon descriptions, and prints ASCII art in the console. Data is stored in Apify datasets. --- ## Features * Scrape web pages using Crawlee + Cheerio * Generate Pokémon details (name, types, stats, description) via Gemini AI * Display Pokémon names in ASCII art in the terminal * Save all Pokémon data to Apify Dataset * Fully configurable via input JSON --- ## Requirements * Node.js 22+ * Apify CLI or running on Apify platform * Environment variable: GEMINI_API_KEY for Google Gemini AI * Optional: Configure input.json to specify number of Pokémon to generate (pokemonCount) --- ## Setup 1. Clone the project bash cd my-actor 2. Install dependencies bash npm install 3. Set environment variable bash export GEMINI_API_KEY="your_google_gemini_api_key_here" 4. Configure input Create an input.json file (optional, defaults provided): json { "pokemonCount": 3 } --- ## Usage ### Run Locally bash node src/main.js ### Run on Apify Platform 1. Upload actor code to Apify. 2. Set the GEMINI_API_KEY in actor settings. 3. Provide input JSON when running the actor. 4. Run the actor; results will be saved to the dataset. --- ## Output * Console: * Pokémon ASCII art * Types, stats, and description printed * Dataset: * JSON objects with all Pokémon data * Field sprite_file stores image filename Example dataset entry: json { "name": "Aurorix", "types": ["Electric", "Ice"], "stats": { "HP": 80, "Attack": 65, "Defense": 75, "Speed": 115, "Special Attack": 120, "Special Defense": 95 }, "description": "Aurorix is said to be born from cosmic dust...", "sprite_file": "Aurorix.png" } --- ## Notes * Gemini AI: Needed for Pokémon generation. Without a valid GEMINI_API_KEY, the AI portion will fail. * ASCII art is generated using Figlet. * Images can optionally be generated and stored if enabled in code. --- ## License MIT License --- This actor is ideal for experimenting with AI-generated content and web scraping combined in a single actor.