Rajasimhareddybolla/SentinelForge ? reverse-engineered prompt

Build me a Python project that acts like an automated AI security testing lab for a chatbot.

I want one command line app that can start a local FastAPI retail chatbot in the background, then run three kinds of checks against it, a multi turn red team attack flow, a broad vulnerability scan, and a fast regression test run. It should save the raw outputs from each run, then combine and clean up the results into one simple Markdown report that groups duplicates and highlights the most important findings first.

Please include a small vulnerable demo chatbot with fake retail logic and some hidden synthetic PII so the tests have something realistic to probe. I also want clear setup instructions, an example environment file, and a clean structure so I can run the scans separately or generate the final correlated report. If you need to look up current docs for the tools involved, feel free to do that.