Appending Observations to History

Evaluating and Debugging Agent Behavior

Real-World Agent Use Cases

Cost, Latency, Reliability Trade-offs

Agent Safety and Permission Scoping

Human-in-the-Loop and Approval Gates

Agent Loop Termination Conditions

Handling Tool Errors and Retries

Agent Frameworks Overview

Multi-Agent Systems and Collaboration

Common Agent Tool Types

Agent Memory: Short-Term vs Long-Term

Planning and Task Decomposition by Agents

Chain-of-Thought Reasoning in Agents

ReAct: Reasoning Meets Acting

Parsing and Executing Tool Calls

How LLMs Choose a Tool

Tool Schemas and Descriptions

Tools and Function Calling Fundamentals

The Agent Loop Explained

Agents vs Plain LLM Prompting

AI Agents: The Sense-Think-Act Loop

Cost and Latency of Agentic Loops

Prompt Injection Threats via Tools

Safety Risks of Autonomous Agents

Human-in-the-Loop Approval

Code Execution and Sandboxing

Multi-Agent Systems and Collaboration

Stopping Conditions and Loop Termination

Multi-Step Reasoning and Iterative Loops

Tool Schema and Argument Definition

Planning and Task Decomposition

Parsing and Validating Tool Outputs

Tool Use Concept and Motivation

Autonomy Levels in Agents

Perception-Reasoning-Action Loop

Agent vs Plain LLM Call

What Defines an AI Agent

Web Browsing and External API Access

Short-Term vs Long-Term Memory

Evaluating Agent Performance

Retrieval as a Tool

Real-World Agent Applications

Error Handling and Recovery

Aggregating Tool Results

Loop Detection Guard

Dangerous Tool Argument Guardrails

Sequential Task Decomposition

Passing Prior Step Output Forward

Coercing Tool Arguments to Numbers

Parsing Action vs Final Answer

Routing to the Correct Tool

Unknown Tool Name Handling

Catching Tool Errors

Parsing Tool Arguments

Validating Required Args

Scratchpad State Scope

Detecting Final Answer

Agent Loop Order

Dispatch Tool By Name

Step Limit Guard

Real-World Applications of AI Agents

Cost and Latency of Agentic Workflows

Evaluating Agent Performance

Code Execution / Code Interpreter Agents

Parsing and Validating Tool Outputs

Multi-Agent Systems and Collaboration

Planning and Task Decomposition

Agent Memory: Short-term vs Long-term

Agent Orchestration Frameworks

Types of Tools

Tool Selection and Routing

Agent vs Standalone LLM Call

Agent Failure Modes and Error Recovery

Stopping Criteria and Loop Termination

Reflection and Self-Correction

Tool Schema and Function Definitions