# Praxis Context
Context management and summarization for AI agent conversations.
## Features
- Token counting with tiktoken
- Automatic summarization
- Context window strategies
- Template-based system prompts
## Usage
```rust
use praxis_context::{DefaultContextStrategy, ContextStrategy};
use praxis_persist::PersistenceClient;
use std::sync::Arc;
let strategy = DefaultContextStrategy::new(
llm_client,
max_tokens: 8000,
);
let context = strategy.get_context_window(
thread_id,
persist_client,
).await?;
```