The Machine Learning Security Atlas
Awesome MLSecOps — Curated ML & AI Security Tools and Resources
MLSecOps (Machine Learning Security Operations) integrates security engineering, threat modeling, testing, supply-chain controls, monitoring, and incident response across the machine-learning lifecycle. It protects data, models, pipelines, infrastructure, LLM applications, and AI agents against poisoning, adversarial manipulation, unsafe artifacts, privacy leakage, model extraction, prompt injection, and excessive agency.
Browse by security problem
Focused guides, one community source
Browse MLSecOps tools and resources by machine learning security category. The GitHub repository contains the complete, community-maintained catalog.
LLM Security and Red Teaming
Test prompts, model behavior, guardrails, and application controls against abuse.
13 catalog entriesModel Scanning and Validation
Inspect model files, notebooks, code, and behavior before release or deployment.
19 catalog entriesAdversarial Machine Learning
Evaluate evasion, poisoning, extraction, inversion, and model robustness.
6 catalog entriesAI Supply-Chain Security
Protect model provenance, artifacts, dependencies, signing, and delivery pipelines.
5 catalog entriesAI Agent and MCP Security
Secure agent tools, memory, identity, permissions, sandboxes, and MCP servers.
12 catalog entriesPrivacy-Preserving Machine Learning
Reduce sensitive-data exposure and test privacy leakage in ML systems.