Project Info

re:zero

This project did not submit a demo video on Devpost.

Inspiration

A week ago, Anthropic released Claude Opus 4.6 after disclosing that it found 500+ vulnerabilities in many popular, well-tested open-source repos/software. We wanted to extend this to EVERY target possible, so that we can improve security faster than agents degrade it.

What it does

It's a platform with an agent named Rem who will reverse-engineer and penetration test anything. For websites, this includes your public site and also authed routes if you give Rem test credentials. She'll even handle 2FA codes by asking you for them when she hits that snag! We used Browserbase/Stagehand for this. For OSS repos, we clone the repo to a Modal sandbox and have Rem run a series of vulnerability tests and just manual exploration of the code to analyze attack surfaces. For hardware, we also use sandboxes to give Rem a scratchpad with which to interact with hardware like drones, webcams, ESP32s, etc. She also has access to an FPGA with a program which can analyze power fluctuations in a device running AES-256 encryption (government comms, financial systems, cloud storage, VPNs, smartphones, etc.) and decrypt their key by measuring the fluctuations in power on the device running encryption.

How we built it

We reverse-engineered the drone by tapping into its wireless protocol and emulating its app's signals to take control of the drone's flight. We also reverse-engineered a webcam by modifying its drivers. For inference and agent sandboxes, we used Modal GPUs, Sandboxes, volumes, and services. All our RL environments use prime-rl environments from Prime Intellect's Environments Hub, and we specifically focused on CTFs, phishing detection, vulnerability detection, and code security. We served our custom-trained models using vLLM running on Modal.

Challenges we ran into

Reverse engineering is not universal, as hardware contains different entry points and different protocols that does not have to be standardized. It was difficult to build a system for distributed reverse engineering. Training the model was also challenging because we had to build and train on our own dataset of phishing scams and GitHub bugs and vulnerabilities across 5 environments. FPGAs are impossibly hard to work with, and our drone has a stabilization problem which we did manage to fix. Tying everything together in a cohesive platform was marvelous but deathly hard because of the really different stacks different targets use. Also, we ran reinforcement learning on an open Nvidia model (OpenReasoning-Nemotron-14b) on cybersec and pentesting, and that was our first pass on a pretty large rollout of GRPO on any model. We used Prime Intellect's (https://primeintellect.ai) stack for RL and verifiers, and Modal for compute.

Accomplishments we're proud of

Rem was able to reverse engineer a drone and a webcam Kenneth and Rem cracked AES-256 encryption, which is one of the most widely-used encryption protocols in the world Rem found several hundred zero-days in OSS repos which we are disclosing to their maintainers directly Rem also found at least 22 REALLY bad exploits on Shresht's personal website (https://tetraslam.world) which he will be fixing right after this hackathon, or so he swears

What we learned

How firmware is implemented and used in conjunction with software. IoT device safety is at greater risk because of AI, even going down to electrical signals. One could steal encryption keys by tapping into a chip's voltage line and monitoring timing fluctuations going through FPGA logic gates. Hardware is cool but agents reverse engineering hardware is even cooler.

What's next

We have everything set up but payments. We only have $200 in Claude API credits remaining so we'll probably set up a Stripe account and see how people like this, especially since it seems highly useful for indie devs and small-scale usage with outsized impact on their security. We'll also be open-sourcing several third-party SDKs we've now created to interface with the variety of hardware that Rem was able to reverse-engineer.

Re:Zero

Reverse Engineer Anything.

The universal autonomous red teaming platform. Point it at any attack surface — source code, web infrastructure, physical hardware — and AI agents will reverse engineer it, find vulnerabilities, and generate a full security audit. One platform, any target, zero prior knowledge required.

The Pitch

There are ~500k unfilled cybersecurity jobs globally. Pen testers cost $200-400/hr. Most companies test annually if at all. Hardware supply chain audits require specialists that barely exist.

Every existing player is vertical-locked: Pentera does networks, Synack (YC W13) does web apps, Snyk does code. Nobody does hardware. But the reasoning pattern is the same across all surfaces:

Recon — map the attack surface
Hypothesize — what could be vulnerable?
Probe — test the hypothesis
Escalate — how far can you go?
Report — document findings + remediation

The only thing that changes is the tool suite. That's Re:Zero.

Stuff we need to do

Get Claude Code SDK on reverse engineering
RL-train GLM-4.7V and Nemotron 3 Nano (30b-A3B) on CTF environments (Prime Intellect environments hub) for reverse engineering and pentesting
Set up a web app (dashboard, "projects", project types so web, codebase, and hardware, with overlap allowed)
Convex backend (including functions which connect to modal and monitor your relevant sandboxes)
Modal sandboxes
Agent harness: skills, ghidra/other mcps, subagents -> opencode?
Targets for demo: open source repos (target C/C++, IoT, MCPs, OpenClaw), drone, 3d printer, laptop
FPGA logic analyzer firmware as a tool for the agent

The Four Modes

Code mode (OSS)
Hardware RE (drone)
Side-channel mode (extract AES key from power traces via FPGA)
Web

Team

Shresht: claude code sdk, agent harnesses, modal sandboxes, convex backend, frontend (live agent streams, camera feed views, waveform visualization for fpga travces, scan results, unified reports), 0-day farming pipeline (on sandboxes), perplexity sonar, demo
Kenneth: ChipWhisperer HDL -> Genesys 2, Pmod ADC, trigger logic, DDR3 trace buffer, python api for agent integration, ESP32 AES target setup with shunt resistor
Mouad: RL training pipeline (Prime Intellect CTF environments, cybersec reasoning, vulnerability reasoning, PCBs, protocol analysis) on GLM-4.7V, Nemotron 3 Nano, create vLLM config for Modal deployment
William: ESP32 probe controller firmware (drone, etc), drone flight controller interface, probe jig, wiring, camera, help shresht with frontend and agent harnesses, join kenneth on fpga

Project Structure

re-zero/
├── web/           # Next.js dashboard, Convex, Clerk auth
├── server/        # FastAPI, Claude Code SDK agents, Modal sandboxes, Perplexity Sonar
├── hardware/      # ESP32 firmware (William), FPGA RTL + Python API (Kenneth)
├── training/      # RL training (Prime Intellect CTF envs), vLLM deploy configs (Mouad)
└── targets/       # demo target info, intentionally vulnerable test apps

Stack

Frontend: Next.js, shadcn/ui, Vercel AI SDK UI, Clerk, Vercel
Database: Convex (real-time sync for live agent state → dashboard)
API + Agents: FastAPI, Claude Code SDK, Anthropic Agent SDK, Pydantic
Compute: Modal (sandboxes for code targets, GPUs for model serving + training)
RL-trained Models: GLM-4.7V, Nemotron 3 Nano — RL via Prime Intellect CTF environments, served via vLLM on Modal
Research: Perplexity Sonar API (CVE lookup, datasheets, component ID)
Hardware Probing: ESP32 (PlatformIO), USB serial bridge
Side-Channel: Kintex-7 FPGA (Verilog), ChipWhisperer Python analyzer
Deployment: Vercel (frontend), Railway (API server), Modal (sandboxes + gpus)

Target customers

Any app/website/software big enough to need security for its users
IoT companies
Data centers
Cloud providers
Defense/government
Automotive
Medical devices
Power grids, water treatment, manufacturing
Semiconductor companies (chip-level security)
Crypto
Anyone who needs to get compliance certified

Tracks we're targeting

Anthropic Claude Code SDK
Modal (sandboxes and main track)
Nvidia (inference and open models track)
Human Capital (company track)
YC track (Salt Security)
Human Capital Fellowship track
Greylock (multiturn agent track)
Vercel (best use of vercel)
Perplexity (Sonar API track)

Analysis

Compare with all teams

View

Metric

Figures cover GitHub contributors during the hackathon window. A co-authored commit counts in full for each author, so per-member totals add up to more than the whole-team figures.

Technology

Found in codeClaimed only

CIn code
CSSIn code
FastAPIIn code
HTMLIn code
JavaScriptIn code
Next.jsIn code
PythonIn code
ReactIn code
Tailwind CSSIn code
TypeScriptIn code
AnthropicClaimed
VercelClaimed

10 of 12 appear in the indexed code. 2 claimed on Devpost could not be matched to code, which may simply mean the tool leaves no trace in the repository.

AI coding agents

Claude CodeConfig · Commits

Detected from committed agent config files and commit authorship. Absence of a signal is not proof an agent was unused.

Codebase size

Source size

1.5 MB

Source files

236

Counts recognized source files only; vendored directories, binaries and lockfiles are excluded, so this is smaller than the repository on disk.