An agent can reverse-engineer apps from traffic — so why can't it learn your habits?
IndependentCoast3974 · reddit · 2026-10-18
A trending GitHub repo lets an agent point at an app and figure out how it behaves purely from network traffic, with no explanation needed. The author contrasts this with still retyping the same preferences to AI every week — watching how you work should be easier than reverse-engineering software. He asks whether the gap is about trust, privacy, or simply a product nobody has built well, and what you'd actually let it observe.
More from coding & agent
- Author and reviewer were both agents — they went 50 rounds arguing — wateriscoding · 2026-10-18
- When Should a Research Agent Stop Searching? A Founder's Three-Part Stopping Rule — Pitiful-Minute-2818 · 2026-10-18
- Teach Your Coding Agent to Prove It's Done: WorkOS on Verification Loops, CLI Checks and LLM Judges — AI Engineer · 2026-10-18
- Claude-written test harness let 31KB fake archives land on a real backup server — Grimmoner · 2026-10-18
- Matt Pocock's pre-commit hook caps agent-written files at 5,000 tokens — mattpocockuk · 2026-10-18
- Backups silently failed for 10 days: three monitoring blind spots lined up — Grimmoner · 2026-10-18