> ## Documentation Index
> Fetch the complete documentation index at: https://praison.ai/docs/llms.txt
> Use this file to discover all available pages before exploring further.
# Context Management
> Manage conversation context size
Context management controls how much history the agent remembers.
```mermaid theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
graph LR
subgraph "Context Window"
M1[Old] --> M2[...]
M2 --> M3[Recent]
M3 --> C[📦 Context]
M1 -.-> X[✂️ Trimmed]
end
classDef old fill:#94A3B8,stroke:#7C90A0,color:#fff
classDef recent fill:#6366F1,stroke:#7C90A0,color:#fff
classDef context fill:#10B981,stroke:#7C90A0,color:#fff
class M1 old
class M2,M3 recent
class C context
```
## Quick Start
```rust theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
use praisonai::{Agent, MemoryConfig};
let agent = Agent::new()
.name("Assistant")
.instructions("You are a helpful assistant")
.memory_config(MemoryConfig::new().max_messages(50)) // Keep last 50 messages
.build()?;
// Agent will automatically trim old messages
let response = agent.chat("Hello!").await?;
```
```rust theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
use praisonai::{Agent, MemoryConfig};
let config = MemoryConfig::new()
.max_messages(100)
.use_short_term(true);
let agent = Agent::new()
.name("Assistant")
.memory_config(config)
.build()?;
```
***
## Context Strategies
| Strategy | Description |
| -------------- | ----------------------- |
| Sliding window | Keep last N messages |
| Summarization | Compress old messages |
| Selective | Keep important messages |
***
## Related
Memory system
Session persistence