Learn how to stay in control as a software architect in the age of coding agents. 🔎 Discover our new training course: Architecting for Agentic Engineering.

Agentic Software Security

2026-11-23 - 2026-11-24

Agentic Software Security (Online) - Felix Schumacher - German

Prompt injection, tool misuse, and uncontrolled tool interactions create attack vectors when operating and using agent-based systems and LLM-powered applications, leading to consequences such as data exfiltration or unauthorized system access. These are real risks that software architects and developers should understand and address.

In our 2-day training course “Agentic Software Security,” we’ll teach you the fundamentals you need to know.

The training will enable you to understand attack vectors such as direct and indirect prompt injection, insecure tool interactions, supply chain risks (e.g., tool poisoning or rug pulls), and cross-context effects; methodically assess risks during the development and integration of agentic applications; and prevent them in a targeted manner with appropriate security measures.

Agenda

Attack Vectors on Generative AI

Agent-Based Systems and Protocols

Your Trainers

socreatory trainer Felix Schumacher

Felix Schumacher

INNOQ

IT-Security

  • Agentic Software Security
  • OWASP Top Ten in practice
  • Web Security

Felix is a Senior Consultant at INNOQ. He enjoys working with IT security, test-driven development and the operation and further development of existing systems.

All info about training