This chapter provides a comprehensive and practical framework for Agentic AI Red Teaming, designed to equip security professionals and AI developers with the tactical skills needed to assess and secure these advanced systems. Grounded in the 12 critical threat categories identified by the Cloud Security Alliance (CSA) and OWASP AI Exchange, this chapter delivers a deep, tactical analysis of each category, complete with detailed attack scenarios and illustrative code examples. Key areas of focus include exploiting agentic tool use, manipulating goal setting and planning logic, poisoning memory and knowledge bases, and executing attacks against multi-agent orchestration protocols. By providing actionable techniques, real-world case study analysis, and a robust ethical framework, this chapter serves as an practical playbook for proactively identifying and mitigating the novel vulnerabilities inherent in Agentic AI, fostering the development of more secure, resilient, and trustworthy autonomous systems.