Projects

Open-source tools and reference implementations that sit behind the papers and articles. 40+ repositories in total on GitHub.

AegisSwarm-Core
Published by CSA
Zero-trust governance for autonomous multi-agent AI networks. Published by the Cloud Security Alliance and cited externally as a reference implementation.
AI securityDefensive security
GSH Framework
Open source
Governed Security Hunting: agentic AI threat hunting mapped to MITRE ATLAS and NIST CSF 2.0, with a full playbook set and CI.
Defensive securityAI security
AEGIS Academic Integrity Checker
Open source
Local-first MCP server and CLI that checks papers for hallucinated citations, AI-text and plagiarism signals, and venue guideline compliance. Listed on mcprush; on PyPI as aegis-integrity.
AI securityDefensive security
ContextGuard
Open source
Zero-trust middleware for Model Context Protocol security.
AI securityDefensive security
ContextGuard-MCP-PQC
IEEE Xplore
Applies RFC 9794 PQ/T hybrid terminology to MCP across four layers, hardening handshakes with hybrid ML-KEM and ML-DSA against harvest-now-decrypt-later and downgrade attacks.
Post-quantum cryptographyAI security
hrpq-gap-scan
Open source
CLI that audits HRPQ preload and cookie-scoping downgrade gaps in post-quantum HTTPS migrations. Companion to the SC World analysis.
Post-quantum cryptographyOffensive security
ARGUS
IEEE Xplore
Autonomous vulnerability discovery for retrieval-augmented and tool-enabled LLM systems.
Offensive securityAI security
genai-llm-pentest-harness
Open source
Reference implementation for the CSO Online pen-testing guide on GenAI, LLM and RAG applications.
Offensive securityAI security
PRAVAL
Open source
Ten-phase agentic vulnerability-assessment framework.
Offensive securityAI security
TRACE-MAS
IEEE Xplore
Cryptographic framework securing multi-agent LLM pipelines against drift, spoofing and prompt injection.
AI securityDefensive security
CausalNetTwin
Open source
Causal digital-twin framework for network security reasoning.
Defensive securityAI security
CADLP
IEEE Xplore
Context-aware DLP proxy that detects and redacts credentials, PII and proprietary code before prompts reach external LLM APIs.
Privacy engineeringAI security
LaptopAI-Agent
Open source
Fully local, privacy-first autonomous agent built on LangGraph, Ollama, ChromaDB RAG and MCP tools.
Privacy engineeringAI security
MarkSentry
Open source
Zero-trust, local-first document-to-Markdown converter with PII auditing.
Privacy engineeringDefensive security
SybilShield-Core
IEEE Xplore
Composite trust scoring for Sybil attack mitigation in permissionless blockchain networks.
Defensive security
DockerDNA
Open source
Layer-by-layer container security analysis: secrets detection, CIS Docker Benchmark mapping, CycloneDX SBOM and SARIF output.
Defensive securityOffensive security
ArchForge
Open source
System design workbench with 60+ interactive patterns and a real-time capacity calculator.

All repositories on GitHub →