Building Modular AI Agent Features with Pydantic AI Capabilities

A developer building a no-code AI agent platform with Pydantic AI introduced modular Capabilities for adding reusable features to agents. The Capabilities, created using AbstractCapability, allow agents to gain web research via Tavily Search API and email functionality via Resend. A Company Knowledge Capability combining vector search and graph queries was also demonstrated.

If you're building AI Agents with Pydantic AI, understanding Capabilities is invaluable - it's the recommended way to add modular, reusable features to your agents. This tutorial is part of my ongoing Pydantic AI series on YouTube https://youtube.com/@joxiahdev , where I build a full no-code AI agent platform from scratch. A capability in Pydantic AI is a modular unit of behavior that can be passed to an AI agent. A capability can give your agent: Think of it as a plug-and-play feature module - build it once, attach it to any agent. Capabilities are created using the Capability or AbstractCapability class: python from pydantic ai.capabilities import AbstractCapability Using AbstractCapability gives you full control over instructions, tools, and behavior. It's Pydantic AI's recommended pattern if you're building a library or platform on top of the framework. In this tutorial, I build a Research Capability powered by the Tavily Search API, and an Email Capability powered by Resend. Here's the research capability: python from pydantic ai import FunctionToolset from pydantic ai.capabilities import AbstractCapability from dataclasses import dataclass from pydantic ai.common tools.tavily import tavily search tool from settings import settings @dataclass class ResearchCapability AbstractCapability : def get instructions self : return "You can use the Tavily search tool for research" def get toolset self : toolset = FunctionToolset toolset.add tool tavily search tool api key=settings.tavily api key return toolset That's it - get instructions tells the agent what it can do, and get toolset gives it the tools to do it. Attach this to any agent, and it instantly gains web research abilities. I also built a Company Knowledge Capability that combines: This lets an agent answer questions from your company's documents, website content, or any knowledge base with both vector search and relationship-aware graph queries. This post covers the concept — the full video walks through building both capabilities live, plus the RAG/GraphRAG ingestion pipeline, observability with Logfire, and wiring it all into a working agent. 🎥 Watch on YouTube → https://youtu.be/ILHtYme4O60 💻 Full source code → https://github.com/bjoxiah/pydantic-ai-series/tree/no-code-agent While you're there, subscribe for more software and AI related content