A JEV-powered Wireshark classifier accidentally uncovered real backdoors on a home network
multiply_matrix · x · 2026-09-23
While visiting family, a user experimented with a TypeSafe JEV-powered classifier on network packets captured via Wireshark. The "for fun" tool flagged serious threats; the user assumed false positives until a frontier model validated the findings — leading to factory-resetting devices and hardening the network.
The author says JEV-powered cybersecurity is real and plans to open-source the packet analyzer. A quoted reply quips: make it smart and cheap, put it everywhere — Jevons paradox.
More from coding & agent
- optimAIzr: A local-first CLI that audits AI token waste, works with Claude Code and Codex — stichstichstich · 2026-09-23
- Given a 3-hour budget and a one-line prompt, Opus 5.5 produced a full Kowloon horror short itself — rainbird · 2026-09-23
- Coinbase launches first fully custodial x402 integration letting AI agents pay for data with USDC — MurrLincoln · 2026-09-23
- Grok in Tesla builds and ships a live Costco deals website from the driver's seat — kieranklaassen · 2026-09-23
- Autonomous AI Agents Are Now Emailing Humans to Register for mnemos Commons — RileyRalmuto · 2026-09-23
- Gemini Spark users list 7 fixes they want: fewer approvals, Gemini integration, better computer use — abelbanko · 2026-09-23