Supercharge your AI agents with data from the web and beyond. Building the library for superintelligence. 🔥
TypeScript aiai-agentsai-crawlerai-scrapingai-searchcrawlerdata-extractionhtml-to-markdownllmmarkdownscraperscrapingweb-crawlerweb-dataweb-data-extractionweb-scraperweb-scrapingweb-searchwebscraping
Add examples/gpt-4.1-crawler
Aparup Ganguly5dca350a11ca
diff --git a/examples/gpt-4.1-web-crawler/.env.example b/examples/gpt-4.1-web-crawler/.env.examplenew file mode 100644index 000000000..307a65f28--- /dev/null+++ b/examples/gpt-4.1-web-crawler/.env.example@@ -0,0 +1,5 @@+# Firecrawl API key+FIRECRAWL_API_KEY=your_firecrawl_api_key_here++# OpenAI API key+OPENAI_API_KEY=your_openai_api_key_here\ No newline at end of filediff --git a/examples/gpt-4.1-web-crawler/.gitignore b/examples/gpt-4.1-web-crawler/.gitignorenew file mode 100644index 000000000..dbb2de947--- /dev/null+++ b/examples/gpt-4.1-web-crawler/.gitignore@@ -0,0 +1,38 @@+__pycache__/+*.py[cod]+*$py.class+*.so+.Python+env/+build/+develop-eggs/+dist/+downloads/+eggs/+.eggs/+lib/+lib64/+parts/+sdist/+var/+*.egg-info/+.installed.cfg+*.egg++# Virtual environment+venv/+ENV/+.env++# IDE specific files+.idea/+.vscode/+*.swp+*.swo++# Logs+*.log++# OS specific files+.DS_Store+Thumbs.db\ No newline at end of filediff --git a/examples/gpt-4.1-web-crawler/README.md b/examples/gpt-4.1-web-crawler/README.mdnew file mode 100644index 000000000..1cf0e5d7f--- /dev/null+++ b/examples/gpt-4.1-web-crawler/README.md@@ -0,0 +1,82 @@+# GPT-4.1 Web Crawler++A smart web crawler powered by GPT-4.1 that intelligently searches websites to find specific information based on user objectives.++## Features++- Intelligently maps website content using semantic search+- Ranks website pages by relevance to your objective+- Extracts structured information using GPT-4.1+- Returns results in clean JSON format++## Prerequisites++- Python 3.8++- Firecrawl API key+- OpenAI API key (with access to GPT-4.1 models)++## Installation++1. Clone this repository:++ ```+ git clone https://github.com/yourusername/gpt-4.1-web-crawler.git+ cd gpt-4.1-web-crawler+ ```++2. Install the required dependencies:++ ```+ pip install -r requirements.txt+ ```++3. Set up environment variables:+ ```+ cp .env.example .env+ ```+ Then edit the `.env` file and add your API keys.++## Usage++Run the script:++```+python gpt-4.1-web-crawler.py+```++The program will prompt you for:++1. The website URL to crawl+2. Your specific objective (what information you want to find)++Example:++```+Enter the website to crawl: https://example.com+Enter your objective: Find the company's leadership team with their roles and short bios+```++The crawler will then:++1. Map the website+2. Identify the most relevant pages+3. Scrape and analyze those pages+4. Return structured information if the objective is met++## How It Works++1. **Mapping**: The crawler uses Firecrawl to map the website structure and find relevant pages based on search terms derived from your objective.++2. **Ranking**: GPT-4.1 analyzes the URLs to determine which pages are most likely to contain the information you're looking for.++3. **Extraction**: The top pages are scraped and analyzed to extract the specific information requested in your objective.++4. **Results**: If found, the information is returned in a clean, structured JSON format.++## License++[MIT License](LICENSE)++## Contributing++Contributions are welcome! Please feel free to submit a Pull Request.diff --git a/examples/gpt-4.1-web-crawler/gpt-4.1-web-crawler.py b/examples/gpt-4.1-web-crawler/gpt-4.1-web-crawler.pynew file mode 100644index 000000000..a82ac0501--- /dev/null+++ b/examples/gpt-4.1-web-crawler/gpt-4.1-web-crawler.py@@ -0,0 +1,261 @@+import os+from firecrawl import FirecrawlApp+import json+from dotenv import load_dotenv+from openai import OpenAI++# ANSI color codes+class Colors:+ CYAN = '\033[96m'+ YELLOW = '\033[93m'+ GREEN = '\033[92m'+ RED = '\033[91m'+ MAGENTA = '\033[95m'+ BLUE = '\033[94m'+ RESET = '\033[0m'++# Load environment variables+load_dotenv()++# Retrieve API keys from environment variables+firecrawl_api_key = os.getenv("FIRECRAWL_API_KEY")+openai_api_key = os.getenv("OPENAI_API_KEY")++# Initialize the FirecrawlApp and OpenAI client+app = FirecrawlApp(api_key=firecrawl_api_key)+client = OpenAI(api_key=openai_api_key)++# Find the page that most likely contains the objective+def find_relevant_page_via_map(objective, url, app, client):+ try:+ print(f"{Colors.CYAN}Understood. The objective is: {objective}{Colors.RESET}")+ print(f"{Colors.CYAN}Initiating search on the website: {url}{Colors.RESET}")++ map_prompt = f"""+ The map function generates a list of URLs from a website and it accepts a search parameter. Based on the objective of: {objective}, come up with a 1-2 word search parameter that will help us find the information we need. Only respond with 1-2 words nothing else.+ """++ print(f"{Colors.YELLOW}Analyzing objective to determine optimal search parameter...{Colors.RESET}")+ completion = client.chat.completions.create(+ model="gpt-4.1-2025-04-14",+ messages=[+ {+ "role": "user",+ "content": [+ {+ "type": "text",+ "text": map_prompt+ }+ ]+ }+ ]+ )++ map_search_parameter = completion.choices[0].message.content+ print(f"{Colors.GREEN}Optimal search parameter identified: {map_search_parameter}{Colors.RESET}")++ print(f"{Colors.YELLOW}Mapping website using the identified search parameter...{Colors.RESET}")+ map_website = app.map_url(url, params={"search": map_search_parameter})++ # Debug print to see the response structure+ print(f"{Colors.MAGENTA}Debug - Map response structure: {json.dumps(map_website, indent=2)}{Colors.RESET}")++ print(f"{Colors.GREEN}Website mapping completed successfully.{Colors.RESET}")++ # Handle the response based on its structure+ if isinstance(map_website, dict):+ # Assuming the links are in a 'urls' or similar key+ links = map_website.get('urls', []) or map_website.get('links', [])+ elif isinstance(map_website, str):+ try:+ parsed = json.loads(map_website)+ links = parsed.get('urls', []) or parsed.get('links', [])+ except json.JSONDecodeError:+ links = []+ else:+ links = map_website if isinstance(map_website, list) else []++ if not links:+ print(f"{Colors.RED}No links found in map response.{Colors.RESET}")+ return None++ rank_prompt = f"""+ Given this list of URLs and the objective: {objective}+ Analyze each URL and rank the top 3 most relevant ones that are most likely to contain the information we need.+ Return your response as a JSON array with exactly 3 objects, each containing:+ - "url": the full URL+ - "relevance_score": number between 0-100 indicating relevance to objective+ - "reason": brief explanation of why this URL is relevant++ Example output:+ [+ {{+ "url": "https://example.com/about",+ "relevance_score": 95,+ "reason": "Main about page containing company information"+ }},+ {{+ "url": "https://example.com/team",+ "relevance_score": 80,+ "reason": "Team page with leadership details"+ }},+ {{+ "url": "https://example.com/contact",+ "relevance_score": 70,+ "reason": "Contact page with location information"+ }}+ ]++ URLs to analyze:+ {json.dumps(links, indent=2)}+ """++ print(f"{Colors.YELLOW}Ranking URLs by relevance to objective...{Colors.RESET}")+ completion = client.chat.completions.create(+ model="gpt-4.1-2025-04-14",+ messages=[+ {+ "role": "user",+ "content": [+ {+ "type": "text",+ "text": rank_prompt+ }+ ]+ }+ ]+ )++ try:+ ranked_results = json.loads(completion.choices[0].message.content)+ links = [result["url"] for result in ranked_results]++ # Print detailed ranking info+ print(f"{Colors.CYAN}Top 3 ranked URLs:{Colors.RESET}")+ for result in ranked_results:+ print(f"{Colors.GREEN}URL: {result['url']}{Colors.RESET}")+ print(f"{Colors.YELLOW}Relevance Score: {result['relevance_score']}{Colors.RESET}")+ print(f"{Colors.BLUE}Reason: {result['reason']}{Colors.RESET}")+ print("---")++ if not links:+ print(f"{Colors.RED}No relevant links identified.{Colors.RESET}")+ return None++ except (json.JSONDecodeError, KeyError) as e:+ print(f"{Colors.RED}Error parsing ranked results: {str(e)}{Colors.RESET}")+ return None++ print(f"{Colors.GREEN}Located {len(links)} relevant links.{Colors.RESET}")+ return links++ except Exception as e:+ print(f"{Colors.RED}Error encountered during relevant page identification: {str(e)}{Colors.RESET}")+ return None++# Scrape the top 3 pages and see if the objective is met, if so return in json format else return None+def find_objective_in_top_pages(map_website, objective, app, client):+ try:+ # Get top 3 links from the map result+ if not map_website:+ print(f"{Colors.RED}No links found to analyze.{Colors.RESET}")+ return None++ top_links = map_website[:3]+ print(f"{Colors.CYAN}Proceeding to analyze top {len(top_links)} links: {top_links}{Colors.RESET}")++ for link in top_links:+ print(f"{Colors.YELLOW}Initiating scrape of page: {link}{Colors.RESET}")+ # Scrape the page+ scrape_result = app.scrape_url(link, params={'formats': ['markdown']})+ print(f"{Colors.GREEN}Page scraping completed successfully.{Colors.RESET}")+++ # Check if objective is met+ check_prompt = f"""+ Given the following scraped content and objective, determine if the objective is met.+ If it is, extract the relevant information in a simple and concise JSON format. Use only the necessary fields and avoid nested structures if possible.+ If the objective is not met with confidence, respond with 'Objective not met'.++ Objective: {objective}+ Scraped content: {scrape_result['markdown']}++ Remember:+ 1. Only return JSON if you are confident the objective is fully met.+ 2. Keep the JSON structure as simple and flat as possible.+ 3. Do not include any explanations or markdown formatting in your response.+ """++ completion = client.chat.completions.create(+ model="gpt-4.1-2025-04-14",+ messages=[+ {+ "role": "user",+ "content": [+ {+ "type": "text",+ "text": check_prompt+ }+ ]+ }+ ]+ )++ result = completion.choices[0].message.content++ if result != "Objective not met":+ print(f"{Colors.GREEN}Objective potentially fulfilled. Relevant information identified.{Colors.RESET}")+ try:+ # Clean up potential markdown formatting or extra text+ if "```json" in result:+ result = result.split("```json")[1].split("```")[0].strip()+ elif "```" in result:+ result = result.split("```")[1].split("```")[0].strip()++ # Try to find JSON content if there's explanatory text+ if "{" in result and "}" in result:+ start_idx = result.find("{")+ end_idx = result.rfind("}") + 1+ if start_idx >= 0 and end_idx > start_idx:+ result = result[start_idx:end_idx]++ return json.loads(result)+ except json.JSONDecodeError as e:+ print(f"{Colors.RED}Error in parsing response: {str(e)}. Proceeding to next page...{Colors.RESET}")+ # Optionally print the raw response for debugging+ # print(f"{Colors.MAGENTA}Raw response: {result}{Colors.RESET}")+ else:+ print(f"{Colors.YELLOW}Objective not met on this page. Proceeding to next link...{Colors.RESET}")++ print(f"{Colors.RED}All available pages analyzed. Objective not fulfilled in examined content.{Colors.RESET}")+ return None++ except Exception as e:+ print(f"{Colors.RED}Error encountered during page analysis: {str(e)}{Colors.RESET}")+ return None++# Main function to execute the process+def main():+ # Get user input+ url = input(f"{Colors.BLUE}Enter the website to crawl : {Colors.RESET}")+ objective = input(f"{Colors.BLUE}Enter your objective: {Colors.RESET}")++ print(f"{Colors.YELLOW}Initiating web crawling process...{Colors.RESET}")+ # Find the relevant page+ map_website = find_relevant_page_via_map(objective, url, app, client)++ if map_website:+ print(f"{Colors.GREEN}Relevant pages identified. Proceeding with detailed analysis using GPT-4.1...{Colors.RESET}")+ # Find objective in top pages+ result = find_objective_in_top_pages(map_website, objective, app, client)++ if result:+ print(f"{Colors.GREEN}Objective successfully fulfilled. Extracted information :{Colors.RESET}")+ print(f"{Colors.MAGENTA}{json.dumps(result, indent=2)}{Colors.RESET}")+ else:+ print(f"{Colors.RED}Unable to fulfill the objective with the available content.{Colors.RESET}")+ else:+ print(f"{Colors.RED}No relevant pages identified. Consider refining the search parameters or trying a different website.{Colors.RESET}")++if __name__ == "__main__":+ main()\ No newline at end of filediff --git a/examples/gpt-4.1-web-crawler/requirements.txt b/examples/gpt-4.1-web-crawler/requirements.txtnew file mode 100644index 000000000..0710ac17e--- /dev/null+++ b/examples/gpt-4.1-web-crawler/requirements.txt@@ -0,0 +1,3 @@+firecrawl==0.11.0+openai==1.14.0+python-dotenv==1.0.0\ No newline at end of file