Skip to main content

How Context Works

When conversations grow longer, they exceed the model’s context window limit:
1

User

First message from user
2

Cerebrum

Response from Cerebrum
3

User

Another message
4

⚠️ Context window limit

Messages below exceed the limit
5

Cerebrum

❌ Can’t fit
6

User

❌ Can’t fit
To solve this, Cerebrum summarizes older messages to make room for new conversations:
1

📦 Summarized Messages

Older messages compressed into summary
2

Cerebrum

Recent response (with overlap from previous context)
3

User

✅ Fits within limit
4

Cerebrum

✅ Fits within limit

Automatic Summarization

Cerebrum uses a sliding window with overlap approach. The overlap ensures context continuity - when summarizing, Cerebrum preserves context from the previous window so important connections aren’t lost.
Summarization happens automatically in the background. You don’t need to do anything special.

What Gets Preserved

When Cerebrum summarizes context, it prioritizes:

Best Practices

When context seems lost, ask Cerebrum to “look at the Canvas” - it shows the current environment, all services, and their configurations.
The Canvas displays the state of a specific environment. Make sure you’re looking at the right env (dev/staging/prod) when discussing changes.
For complex discussions, provide your own summary: “To recap, we decided to use X because Y.”
For very large projects, consider completing major milestones before moving on. This creates natural breakpoints.

Context Limits by Model

Different models have different context windows:
Larger context windows mean less frequent summarization, but all models benefit from Cerebrum’s smart context management.