Offensive Lab

Offensive security research at the edge — satellite protocols, adversarial AI, and novel attack surfaces.

Overview

An open-source collection of offensive security tools targeting underexplored attack surfaces — satellite communication protocols, AI/ML systems, and the intersection of autonomous systems with space infrastructure. Built for authorized security research, red teaming, and vulnerability discovery.

---

Tools

satcom_fuzzer — CCSDS Space Protocol Fuzzer

AI-guided protocol fuzzer for CCSDS telecommand and telemetry packets. Uses mutation strategies informed by ML to maximize code coverage and discover parsing vulnerabilities in ground station software.

• CCSDS TC/TM packet generation & mutation
• AI-guided fuzzing with coverage feedback loop
• Supports AOS, Proximity-1, and custom frame types
• Crash triage and deduplication
• Export to PCAP for Wireshark analysis

orbital_recon — Satellite OSINT & Attack Surface Mapper

Passive reconnaissance framework for satellite infrastructure. Pulls public TLE data, calculates orbital passes, enumerates known ground station locations, and maps RF attack surfaces.

• TLE fetching & orbital propagation (SGP4)
• Ground station geolocation database
• RF frequency band enumeration per satellite
• Pass prediction for signal interception windows
• Link budget calculator for SDR intercept feasibility

ai_adversarial — Adversarial ML Attack Framework

Toolkit for attacking ML models deployed in autonomous and space systems — satellite imagery classifiers, anomaly detectors, and AI-driven C2 systems.

• Evasion attacks (FGSM, PGD, C&W) against image classifiers
• Data poisoning payload generator
• Model extraction via query API
• Prompt injection payloads for LLM-augmented systems
• Membership inference attacks

space_protocol_analyzer — Space Link Protocol Dissector

Deep packet inspection and vulnerability analysis for space communication protocols.

• CCSDS Space Packet Protocol parser
• Proximity-1 frame dissector
• AOS frame analysis
• Authentication bypass detection
• Command injection vector mapping

ai_adversarial/supply_chain — ML Supply Chain Attacker

Offensive toolkit targeting the ML model distribution pipeline — pickle RCE, ONNX graph injection, safetensors fuzzing, and model registry typosquatting.

• Pickle deserialization RCE payload generator
• ONNX graph backdoor injection
• Safetensors header fuzzing
• Model file static analyzer
• HuggingFace Hub typosquatting templates

satcom_fuzzer/rf_sigint — RF Signal Intelligence & SDR Simulator

Simulates software-defined radio attacks against satellite links.

• BPSK/QPSK IQ sample generation
• Jamming effectiveness calculator (J/S ratio)
• GNSS spoofing parameter generator
• RF fingerprinting for transmitter authentication/spoofing
• Multiple jamming modes: barrage, spot, sweep, pulse

---

Quick Start

git clone https://github.com/Venkatatadu/offensive-lab.git
cd offensive-lab
pip install -r requirements.txt

# Run the CCSDS fuzzer
python -m satcom_fuzzer --target localhost:9999 --mode ai-guided --frames 10000

# Satellite recon
python -m orbital_recon --sat "ISS" --passes 5 --location "46.1,-64.8"

# Adversarial ML
python -m ai_adversarial --attack fgsm --model target_model.onnx --input sample.png

# Protocol analysis
python -m space_protocol_analyzer --pcap capture.pcap --detect-vulns

---

Legal

These tools are for authorized security research only. Usage against systems without explicit written authorization is illegal and unethical. All satellite data used comes from publicly available sources (CelesTrak, Space-Track.org).

Contributing

PRs welcome. See CONTRIBUTING.md.

License

MIT — see LICENSE.