MLSecOps Practical Reference Guide

Open-source MLSecOps handbook for AI security, LLM/RAG, and secure MLOps.

v1.1.0 — practical reference for securing AI systems across the ML lifecycle: data, training, deployment, runtime, SOC, and governance.

GitHub repository · Getting Started · Zenodo DOI

Topics

Area	Start here
MLSecOps lifecycle	Chapter 6 — Lifecycle control model
LLM & RAG security	Chapter 7
Agentic AI & MCP	Chapter 8
AI supply chain	Chapter 5
Implementation rollout	Appendix E

Start reading

Chapter 1 — Introduction	Scope, principles, lifecycle overview
Chapter 6 — Lifecycle control model	Ten control points and release decisions
Appendix E — Implementation Reference	Architecture cards, templates, playbooks
Full table of contents	All sections

What this guide adds

1. Ten lifecycle control points
2. Release decisions separate from evidence-producing steps
3. Evidence Pack per release
4. Unified thread: threat → runtime → SOC → governance

Details in Chapter 1.