Module 04 Agent Builder Knowledge Bases and Guardrails

🏗️ Module 04 · Agent Builder, Knowledge Bases and Guardrails

The low-code surface, and how to wire knowledge and safety into it.

Module 04 of 16 · track Invoke · Estimated time: 4–5 hours · Prerequisites: Module 03.

Agent Builder is the fastest path from idea to working agent. This module shows what it gives you for free, where its ceiling is, and how knowledge bases and guardrails attach to a real agent.

Open the module on GitHub ↗: the slides, notebooks, exercises and solutions are there. This page is the reading order, with the tutorial and the workbench alongside.

Read first, on the site

The lesson that frames this module is AI Agent Guardrails That Hold: Why a Prompt Is Not a Control (7 min): which controls hold and which are only prompts. Then try the authority and gate mapper in the workbench: the same decision, with your numbers.

What you will be able to do

• Build an agent in Agent Builder and understand what it generates
• Attach a knowledge base and verify grounding actually happened
• Configure guardrails and test them against adversarial input
• Know when to graduate from Agent Builder to code

Concepts in this module

Portable GenAI concepts	AWS-specific surface
Low-code agent composition	Bedrock Agent Builder
Retrieval grounding	Bedrock Knowledge Bases
Safety policy design	Bedrock Guardrails
	Lambda action groups
	S3 data sources

Portable concepts transfer to any stack. The AWS column is where this module touches the cloud — see docs/concepts/portability-matrix.md.

Run it in this order

#		Step	What it is
1	📖	slides/agent_builder_deck.md	Agent Builder deck
2	💻	notebooks/agent_builder_notebook.ipynb	Agent Builder walkthrough
3	📖	slides/kb_guardrails_deck.md	Knowledge bases and guardrails
4	💻	notebooks/kb_guardrails_agentbuilder_notebook.ipynb	KB + guardrails in Agent Builder
5	💻	notebooks/kb_guardrails_travelmind_notebook.ipynb	TravelMind with KB and guardrails
6	🔖	guides/TravelMind_ActionGroups_Lambda_RoC_Runbook.md	Action groups + Lambda runbook

📖 read   💻 run   ✏️ practise   📊 workbook   🔖 reference

This module has no separate exercises or solutions. It is a guided build: the three notebooks are the practice, and each is worked end to end with the decisions explained as they are made.

For graded practice on the same material, use Module 03's exercises (agent behaviour and action groups) and Module 10's RAG exercises (retrieval quality), both of which have worked solutions.

Common mistakes

• Assuming a knowledge base is grounded because it returned text. Check the citations.
• Guardrails configured but never adversarially tested.

When it breaks

Every error anyone has hit in this course is in the Error Index, by the exact string. Model access and region problems are in Model and Region Notes. If it is a question rather than an error, check the answered ones first, then ask in Q&A.

Field guide for this module

Frameworks

• Grounding Triangle — Is it retrieved, cited, or actually verified?
• Blast Radius Grid — Score every tool before you wire it

Quick reference

• RAG pipeline — Every stage, and the failure each causes
• IAM for agents — KB service roles and the three-role split

Recipes and procedures

• Runbook · stale knowledge base — The silent failure of a KB nobody re-syncs

---

← Module 03 · Amazon Bedrock Agents · All modules · Module 05 · The Agent Loop: No Framework to Strands →

_{✏️ This page is generated from modules/04-agent-builder-and-knowledge-bases/README.md. To change it, edit the module README: an edit made here is replaced at the next sync.}