Agent Quest

Tracks

1. 01🧠What Are AI Agents?

   0/5 lessons

   Foundations

   The mental model: an agent is a model inside a controlled loop that can use tools, hold state, and work toward a goal.

   Lesson list (5)

   • 01
     Chatbot vs Agent: Text Engine or Action System~22 min · agent-loop, autonomy, tools
   • 02
     The Agent Loop: OODA for Software~24 min · ooda, control-loop, stop-conditions
   • 03
     ReAct and the Agent Landscape~34 min · react, frameworks, landscape
   • 04
     Agent Anatomy: Brain, Tools, Memory, Policy~26 min · architecture, components, policy
   • 05
     Agent or Workflow? Choosing the Right Autonomy~24 min · workflow, autonomy, architecture

   ✓Quiz · 4 questions→

2. 02🔄Single Agent

   0/5 lessons

   The Core Loop

   Build one agent well before building a team of agents. This track covers the raw loop, OpenAI function calling, Claude tool use, and prompt-state contracts.

   Lesson list (5)

   • 01
     The Minimal Agent: Loop Before Framework~32 min · python, minimal-agent, loop
   • 02
     OpenAI Responses API: Function Calling Loop~34 min · openai, responses-api, function-calling
   • 03
     Claude Tool Use: tool_use and tool_result~34 min · anthropic, claude, tool-use
   • 04
     Error Handling and Context Budget~36 min · errors, context, budgets
   • 05
     System Prompt and State Contract~28 min · system-prompt, state, contracts

   ✓Quiz · 4 questions→

3. 03🛠️Tools

   0/5 lessons

   Giving Agents Hands

   Tool quality determines whether an agent can act precisely, safely, and cheaply.

   Lesson list (5)

   • 01
     Tool Schema Design: The Model Reads Your API~30 min · tool-schema, json-schema, descriptions
   • 02
     Tool Results: Small, Structured, Actionable~26 min · tool-results, context-budget, json
   • 03
     Common Production Tools: Files, Web, Code~38 min · files, web, sandbox, code-execution
   • 04
     Permissions, Sandboxes, and Human Approval~30 min · permissions, sandbox, approval
   • 05
     MCP and Connectors: Standardizing the Hands~28 min · mcp, connectors, resources

   ✓Quiz · 4 questions→

4. 04💾Memory

   0/5 lessons

   Agents That Remember

   Memory turns isolated tool calls into continuity: working state, retrieval, structured facts, and compaction.

   Lesson list (5)

   • 01
     Working Memory: Scratchpads and Run State~26 min · scratchpad, state
   • 02
     Memory Architecture: Episodic, Semantic, Procedural~34 min · memory, rag, structured-state
   • 03
     Vector Memory and Retrieval~30 min · vector, rag, embeddings
   • 04
     Structured Memory: Facts, Preferences, Procedures~28 min · sql, facts, preferences
   • 05
     Context Compaction Without Amnesia~30 min · context-window, compaction, summaries

   ✓Quiz · 4 questions→

5. 05🗺️Planning

   0/5 lessons

   Agents That Think Ahead

   Planning is how agents turn vague goals into sequenced work, verification, replanning, and human checkpoints.

   Lesson list (5)

   • 01
     Plan-Execute: Stop Wandering~30 min · planning, plan-execute
   • 02
     Advanced Planning: Branches, Review, Replan~36 min · planning, tree-of-thought, review
   • 03
     Reflection and Verification~28 min · verification, reflection
   • 04
     Failure Handling and Replanning~28 min · failure, replanning
   • 05
     Human-in-the-Loop Checkpoints~26 min · human-in-loop, approval

   ✓Quiz · 4 questions→

6. 06🤝Multi-Agent Systems

   0/5 lessons

   Agents Working Together

   Multi-agent systems add specialization, review, handoffs, and parallelism, but they multiply cost and debugging surface.

   Lesson list (5)

   • 01
     Orchestrator, Pipeline, Debate, Review~30 min · orchestration, patterns
   • 02
     CrewAI and Role-Based Teams~36 min · crewai, roles, teams
   • 03
     Handoffs vs Agents as Tools~28 min · handoffs, delegation
   • 04
     Critique, Debate, and Adversarial Review~28 min · critique, debate, review
   • 05
     Cost, Latency, and Debugging Surface~26 min · cost, latency, debugging

   ✓Quiz · 4 questions→

7. 07📦Agent Frameworks

   0/5 lessons

   The Ecosystem

   Frameworks are useful when their primitives match your problem. This track covers DIY, OpenAI Agents SDK, LangGraph, and Claude Agent SDK.

   Lesson list (5)

   • 01
     DIY or Framework?~26 min · frameworks, architecture
   • 02
     Framework Landscape and Decision Matrix~36 min · frameworks, decision-matrix
   • 03
     OpenAI Agents SDK: Agents, Tools, Handoffs, Guardrails~32 min · openai-agents-sdk, guardrails
   • 04
     LangGraph: Durable Stateful Workflows~32 min · langgraph, stategraph
   • 05
     Claude Agent SDK and MCP-Native Work~30 min · claude-sdk, mcp, coding-agents

   ✓Quiz · 4 questions→

8. 08🚀Production Agents

   0/5 lessons

   Reliability and Deployment

   Production agents need limits, observability, evaluation, security, deployment shape, and graceful failure.

   Lesson list (5)

   • 01
     Reliability Limits: Steps, Time, Cost, Context~30 min · reliability, limits
   • 02
     Observability and Evaluation~32 min · observability, evals
   • 03
     Testing, Security, and Deployment Patterns~40 min · testing, security, deployment
   • 04
     Prompt Injection and Tool Security~32 min · security, prompt-injection
   • 05
     Deployment Patterns and Graceful Failure~28 min · deployment, queues, fallback

   ✓Quiz · 4 questions→