Struct LLMCompactor 

pub struct LLMCompactor { /* private fields */ }

LLM-based compactor that summarizes older conversation using an LLM. It replaces older messages with a single summary message while preserving recent turns.

Implementations

impl LLMCompactor

pub fn new( client: LLMClient, config: LLMCompactorConfig, ) -> Result<Self, CompactorConfigError>

Creates a new LLM compactor.

Arguments

• client - LLMClient client for making LLM calls
• config - Configuration for compaction behavior

Returns

Error if configuration is invalid.

Trait Implementations

impl AsyncCompactor for LLMCompactor

fn compact_async<'a>( &'a self, conversation: Vec<Message>, _compact_summaries: &'a HashMap<String, String>, ) -> Pin<Box<dyn Future<Output = Result<(Vec<Message>, CompactionResult), CompactionError>> + Send + 'a>>

Performs async compaction that may involve LLM calls. Returns the new conversation and compaction statistics.

impl Compactor for LLMCompactor

fn should_compact(&self, context_used: i64, context_limit: i32) -> bool

Returns true if compaction should occur before this LLM call.

fn compact( &self, _conversation: &mut Vec<Message>, _compact_summaries: &HashMap<String, String>, ) -> CompactionResult

Performs synchronous compaction on the conversation. Returns the compaction statistics.

fn is_async(&self) -> bool

Returns true if this compactor requires async compaction. When true, the caller should use AsyncCompactor::compact_async() instead of compact().