LLM assisted reverse engineeringImage: slowchronicle.com
Cooling

LLM assisted reverse engineering

Heat 42 · #46 in AI · Cooling2 reports · 2 sources · First reported Nov 18 · Updated Nov 20

The story

A reverse engineer with 20 years of experience hacking hardware and software describes switching from manual work to an LLM-agent workflow, and claims tasks that would take years each by hand are now succeeding within six months. The setup pairs a DeepSeek v4.1 Flash harness with a Ghidra MCP bridge, radare2 and dedicated agent skills. His claimed targets include the Alphasmart NEO, the EmWave biofeedback device and Sony Alpha camera protocols. The results are the author's own claims; no second source has covered the post.
Latest development
enters on single-platform composite (IceWM/Switcher/Smart-people precedent).

How it unfolded

  1. enters on single-platform composite (IceWM/Switcher/Smart-people precedent).
  2. claims 6 months of AI-assisted successes that would take years each by hand (Alphasmart NEO, EmWave biofeedback device, Sony Alpha camera protocols). Score 68.0 (rank 85, discussion 70, cross 30: HN only). Search review this round: no second source covering this specific post (other hits are generic LLM+RE papers)
  3. author (20 years hacking hardware and software, previously manual) describes an LLM-agent workflow: DeepSeek v4.1 Flash harness + Ghidra MCP bridge, radare2/kuna, dedicated agent skills

The angle

How one reverser points an LLM agent at a binary and lets it do the disassembly: the exact tool stack and what it actually crackedSuggested format: Short video · for Security researchers, reverse engineers, AI tooling audience

Coverage (2)

Heat

Now 42 · Peak 42 · 64 recorded checks since Nov 20
Full ranking →