Discover Hidden Website API Endpoints Using Regex and AI

This n8n workflow automates the discovery of hidden or undocumented API endpoints on websites lacking public APIs. It fetches HTML from a target site, extracts embedded JavaScript files, and applies two complementary methods: a generic regex to capture all potential URLs, and AI-driven analysis where the first LLM identifies endpoints, methods, and parameters, while a second AI agent generates tailored regex patterns per JS file.

Key features include AI-supported endpoint extraction with detailed descriptions, custom regex generation, iterative regex validation via an AI agent tool, and side-by-side comparison of regex vs. AI results. This ensures comprehensive coverage without missing critical endpoints and refines accuracy through validation loops.

Ideal for developers building automations for platforms without official APIs, it saves time over manual inspection or scraping. Use cases include reverse-engineering web apps for integrations, prototyping API clients, or auditing internal calls in SPAs. Benefits: high precision via hybrid regex-AI, reduces trial-and-error, and scales to multiple sites effortlessly.