Comprehensive security skills for agentic coding based on OWASP Agentic AI Security Initiative (ASI) 2026 guidelines. Use this skill whenever building, reviewing, or modifying agentic systems, multi-agent architectures, LLM-integrated applications, tool-calling pipelines, or any code where an AI agent acts autonomously on behalf of a user. Covers goal integrity, secure tooling, supply chain safety, memory hygiene, inter-agent coordination, and human-agent safeguards.