Struct CompressionConfig 

pub struct CompressionConfig {
    pub strategy: CompressionStrategy,
    pub pruning_strategy: PruningStrategy,
    pub model: String,
    pub compress_provider: ProviderName,
    pub probe: CompactionProbeConfig,
    pub archive_tool_outputs: bool,
    pub focus_scorer_provider: ProviderName,
    pub high_density_budget: f32,
    pub low_density_budget: f32,
    pub predictor: CompressionPredictorConfig,
}

Configuration for active context compression (#1161).

Fields

strategy: CompressionStrategy

Compression strategy.

pruning_strategy: PruningStrategy

Tool-output pruning strategy (requires context-compression feature).

model: String

Model to use for compression summaries.

Currently unused — the primary summary provider is used regardless of this value. Reserved for future per-compression model selection. Setting this field has no effect.

compress_provider: ProviderName

Provider name from [[llm.providers]] for compress_context summaries. Falls back to the primary provider when empty. Default: "".

probe: CompactionProbeConfig

Compaction probe: validates summary quality before committing it (#1609).

archive_tool_outputs: bool

Archive tool output bodies to SQLite before compaction (Memex #2432).

When enabled, tool output bodies in the compaction range are saved to tool_overflow with archive_type = 'archive' before summarization. The LLM summarizes placeholder messages; archived content is appended as a postfix after summarization so references survive compaction. Default: false.

focus_scorer_provider: ProviderName

Provider for Focus strategy segment scoring (#2510). Falls back to the primary provider when empty. Default: "".

high_density_budget: f32

Token-budget fraction for high-density content in density-aware compression (#2481). Must sum to 1.0 with low_density_budget. Default: 0.7.

low_density_budget: f32

Token-budget fraction for low-density content in density-aware compression (#2481). Must sum to 1.0 with high_density_budget. Default: 0.3.

predictor: CompressionPredictorConfig

Performance-floor compression ratio predictor (#2460).

Trait Implementations

impl Clone for CompressionConfig

fn clone(&self) -> CompressionConfig

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for CompressionConfig

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for CompressionConfig

fn default() -> CompressionConfig

Returns the “default value” for a type.

impl<'de> Deserialize<'de> for CompressionConfig

where CompressionConfig: Default,

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for CompressionConfig

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.