Back to Workshops
Workshop Free with sign-in

Building Safe AI Agents with Guardrails

January 6, 2026 Alexey Grigorev
ai-agents llm-engineering agent-safety tooling-architecture async-control

We start with a DataTalks.Club Data Engineering Zoomcamp FAQ assistant. Then we add checks that keep the agent on topic and block unsafe responses. The checks also show how to cancel wasted work when a guardrail fails. We first use the OpenAI Agents SDK for its built-in guardrails. Then we rebuild the same idea with tools and plain asyncio, so it works with other agent frameworks.

Links

The external resources:

The notebook you will build

By the end, you wrap a tool-using FAQ agent in guardrails:

flowchart LR USER["User question"] INPUT["Input topic guardrail"] FAQ["FAQ assistant agent"] SEARCH["search_faq tool minsearch index"] OUTPUT["Output safety guardrail"] ANSWER["User-facing answer"] OPENAI["OpenAI model calls"] USER --> INPUT INPUT -->|passes| FAQ INPUT -->|trips| ANSWER FAQ --> SEARCH FAQ --> OPENAI INPUT --> OPENAI OUTPUT --> OPENAI FAQ --> OUTPUT OUTPUT -->|passes| ANSWER OUTPUT -->|trips| ANSWER

The base agent can already search the FAQ, but it tries to answer unrelated questions too. The input guardrail blocks questions outside the course domain. The output guardrail checks the agent response for policy problems. Examples are promising deadline extensions or writing homework for a student.

In the later parts we show the same checks as tools. We also build a small async runner that can cancel work when a guardrail trips.

Free with sign-in

Sign in to access this workshop

Sign up with Google Sign up with GitHub
Create a free account Already have an account? Sign in

Tutorial pages

  1. Overview and setup
  2. Part 1: Load FAQ documents
  3. Part 2: Base FAQ agent
  4. Part 3: Input guardrail
  5. Part 4: Tripwire handling
  6. Part 5: Output guardrail
  7. Part 6: Multiple guardrails
  8. Part 7: Streaming with guardrails
  9. Part 8: Tool-based guardrails
  10. Q&A: side discussions
  11. Part 9: Async primer
  12. Part 10: Cancel on tripwire
  13. Part 11: DIY runner for the FAQ agent
  14. Summary and next steps

Watch the recording

Recording available with Free with sign-in.

Free with sign-in View code on GitHub

Related content

Event Free with sign-in

Building Safe AI Agents with Guardrails

We start with a DataTalks.Club Data Engineering Zoomcamp FAQ assistant. Then we add checks that keep the agent on topic and block unsafe responses. The checks also show how to can…

January 6, 2026

ai-agents llm-engineering
Workshop Free with sign-in

Build a Documentation RAG Agent with Search Tools

We build a RAG system over real documentation and evolve it into an agentic search flow. For the knowledge base we use the Evidently AI documentation, a real set of Markdown files…

May 04, 2026

rag ai-agents
Workshop Free with sign-in

Build a Coding Agent with Tools, Skills and PydanticAI

We build a coding agent from scratch, starting with tool calls and an agentic loop. We add a skills system that loads reusable instruction files on demand, and the final version r…

April 21, 2026

ai-agents llm-engineering