---
published: true
id: agentic-capabilities
title: Agentic Capabilities
description: Knowledge captured in João Lopes' learning graph.
tags:
- ai-engineering
- agent-design
locale: en
resource: https://joaolopes.dev.br/knowledge/concepts/en/agentic-capabilities/
created: '2026-04-05'
---

# Agentic Capabilities
The configurable building blocks of an AI coding agent: instructions, rules, memory, tools, skills, and hooks.

## Definition
Agentic capabilities are the mechanisms you configure to shape how an agent behaves. Instructions and rules load passive context. Memory persists learnings across sessions. Tools give the agent functions to call. Skills define reusable workflows. Hooks intercept tool calls and enforce constraints. Each capability shapes what the model sees in the context window or how the harness executes around it.

## Source
- [How Claude Code works](https://code.claude.com/docs/en/how-claude-code-works)
- [Best practices for coding with agents](https://cursor.com/blog/agent-best-practices)

## Related
- [Agentic Harness](/concepts/en/agentic-harness.md)
- [Context Window](/concepts/en/context-window.md)
- [Instructions](/concepts/en/instructions.md)