Context Engineering

Most agent problems trace back to context. This skill gives you a framework for feeding Claude the right information at the right time: rules files for project conventions, selective spec loading instead of dumping everything, relevant source files before editing, and clean error messages when things break. The hierarchy diagram alone (rules → specs → source → errors → history) is worth internalizing. Best part is the confusion management section, which teaches the agent to surface conflicts instead of guessing. Use this when starting a project, when output quality drops, or when the agent starts hallucinating APIs that don't exist.

skills.sh installs: 71.1K
GitHub stars: 114.6K
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add addyosmani/agent-skills --skill context-engineering --agent claude-code

Installs into .claude/skills of the current project.

Files

Context Engineering

Overview

Feed agents the right information at the right time. Context is the single biggest lever for agent output quality — too little and the agent hallucinates, too much and it loses focus. Context engineering is the practice of deliberately curating what the agent sees, when it sees it, and how it's structured.

When to Use

Starting a new coding session
Agent output quality is declining (wrong patterns, hallucinated APIs, ignoring conventions)
Switching between different parts of a codebase
Setting up a new project for AI-assisted development
The agent is not following project conventions