use gcp’s free 300 usd credits to bulk-scan code for patterns that standard sast tools WILL miss. once you have a list of potential vulnerabilities, feed them into chatgpt o1 for deeper analysis and refined insights. this combined approach leverages automated scale with llm-based intelligence, making it easier than ever to scrape low-hanging fruit in open-source projects like gitlab. proof of concept for gitlab’s program coming soon.