agentic-patterns

What problem does it solve?

This Skill provides a comprehensive guide to designing, building, and securing AI agents, addressing common pitfalls like infinite loops, prompt injection, and hallucinations.

Core Features & Use Cases

• Agent Loop Design: Understand the PERCEIVE-PLAN-ACT-OBSERVE cycle and termination conditions.
• Tool Calling: Securely design and execute tools for LLMs.
• Memory Architectures: Implement various memory types for effective context management.
• Coordination Patterns: Learn supervisor and peer review models for multi-agent systems.
• Human-in-the-Loop: Integrate essential human oversight for critical actions.
• Guardrails: Implement input, output, and resource safety checks.
• Use Case: When developing a new autonomous AI agent, consult this Skill to ensure it incorporates best practices for safety, reliability, and efficient operation.

Quick Start

Use the agentic-patterns skill to understand how to implement human-in-the-loop gates in AI agents.