ni5arga/deanonymizer ? reverse-engineered prompt

Build me a command line tool that can take one or more public usernames and tell me how much they accidentally reveal about themselves online.

It should look at public activity from places like Reddit, Hacker News, GitHub, Stack Overflow, and any personal website linked from those profiles, then pull out clues like emails, cross platform handles, location hints, work history, routines, and other identifying details. I want it to combine those clues into a readable risk report with clear confidence levels, the exact evidence for each claim, and a short suggestion for how to reduce the exposure. Please make sure the report is honest about uncertainty and includes a section for direct identifiers like emails and linked accounts near the top.

I’d also like JSON output for automation, plus a strict mode that fails if there is no proof from outside the profile itself. If you need to, look up the current docs for the APIs and model providers online and make sensible choices.