What Context Actually Is
What you learn here
Context is all the information you give Claude to help it understand the task correctly. But context is not the same as everything you know about something. It is what Claude needs right now.
Key principles
- Relevant is not the same as necessary. Include only what directly enables correct results.
- More is not always better. Excess context introduces noise and reduces clarity.
The mental model
Think about context as signal-to-noise optimization. Your job is to maximize signal so Claude sees what matters, and minimize noise so it does not drift.
Context is a tool for accuracy, not a container for knowledge. Provide only what the task requires at this moment.
Key Takeaways
- Context is the information Claude needs to complete the task
- Necessary beats relevant every time
- Less noise means clearer reasoning