
This is the conceptual foundation for the entire context engineering collection.
$ npx -y skills add muratcankoylan/agent-skills-for-context-engineering --skill context-fundamentals --agent claude-codeInstalls into .claude/skills of the current project.
Sponsors
It explains what context actually is (system prompts, tool definitions, retrieved documents, message history, tool outputs), how attention mechanics constrain agent behavior (the U-shaped attention curve, the n-squared relationship problem), and why quality beats quantity when you're working against a finite budget. Use it for onboarding, design decisions, and grounding operational work in first principles. It deliberately routes execution elsewhere: debugging attention failures to context-degradation, token efficiency to context-optimization, conversation summarization to context-compression. The four core principles (informativity over exhaustiveness, position-aware placement, progressive disclosure, iterative curation) are the mental models you need before touching the specialized skills.
Select a file.