Building the inference layer as the primary enterprise operating layer.
"We build open-source tools at the intersection of AI agents, cybersecurity, and local inference — with a bias toward production-grade security, Apple Silicon optimization, and agentic automation."
| Repo | What it does |
|---|---|
| TSLIT | Time-Shift LLM Integrity Tester — 3,840-prompt adversarial evaluation harness for local LLMs |
| pentest-regulatory-intel | AI pentesting regulatory intelligence across 20+ financial jurisdictions |
| strixresearch | Research docs for Strix autonomous pentesting platform |
| agentic-ai-pentesting | Agentic AI for Layer 7 penetration testing |
| Repo | What it does |
|---|---|
| oscal-agent-guardrails | OSCAL controls as policy guardrails for LangGraph agents |
| oscal-digital-twin-playground | OSCAL-backed digital twin with risk assessment agents |
| oscal-zero-trust-lens | Zero Trust semantic overlay on OSCAL controls |
| oscal-agent-lab | Multi-agent copilot for OSCAL catalogs and SSPs |
| oscal-cac-playgd | Compliance-as-code CLI using real NIST OSCAL JSON |
| Repo | What it does |
|---|---|
| agent-stack | Interactive 10-layer Agent Stack architecture visualization |
| deepagent-azure-cli | LangChain DeepAgents + Azure OpenAI coding assistant CLI |
| N8n2langraph | Convert n8n workflows to standalone LangGraph scripts |
| sst-autoresearch | Autonomous research pipeline |
| projectpulse | SAP meetings → knowledge base + podcast via LangGraph |
| Repo | What it does |
|---|---|
| tars-ai | Talk to TARS — fully local voice agent on MLX, no API keys |
| screenlens | Local video scene intelligence for Apple Silicon |
| qwenbench-mlx | Benchmark the full Qwen 3.5 family (0.8B–35B) on Apple Silicon |
| mlx-distillation-explained | Educational: model distillation with Claude teacher + MLX student |
| local-mlx-responsesAPI-server | OpenAI-compatible local inference server via MLX |
| audiobook_generator | Books → audiobooks with Qwen3-TTS + LangGraph |
| QWEN3-VL-Python-OCR-Script-MLX | Image captioning with local MLX vision-language models |
| MLX-YouTubeScribe | Audio/video transcription with WhisperX |
| Repo | What it does |
|---|---|
| gitnexus_fleet | Clone, index, and query your entire GitHub fleet via MCP |
AI Agents in Cybersecurity — A Practitioner's Guide, from Strategy to Implementation.
How autonomous agents are transforming enterprise security — from automating SOC operations and incident triage to building multi-agent defense systems. Includes real-world case studies, practical frameworks, and a hands-on lab companion.
Built by Nic Cravino · AI/ML Engineering · Cybersecurity · Enterprise Automation