Struct ContextConfig 

pub struct ContextConfig {
    pub manager: ContextManager,
    pub features: ContextFeatures,
    pub semantic: SemanticConfig,
    pub estimation: TokenEstimation,
    pub summary_input_cap_floor_chars: usize,
    pub api_surface: ApiSurfaceConfig,
    pub input_ceiling_pct: u32,
    pub low_budget_pct: usize,
}

[context] config section (Step 24.8, #559; features added Phase 26, #588).

Fields

manager: ContextManager

Context-management strategy preset. Default standard.

features: ContextFeatures

Per-feature overrides under [context.features] — each inherits the manager preset’s default unless explicitly set (Phase 26, #588).

semantic: SemanticConfig

[context.semantic] — settings for the semantic RAG feature (Step 26.5, #582).

estimation: TokenEstimation

[context.estimation] — the cheap token-estimation heuristic (chars_per_token, default 4). Threaded through the estimators; the per-model calibration ratio scales the result on top.

summary_input_cap_floor_chars: usize

Floor (chars) for the whole-middle summarizer input cap. The cap is normally the compression budget converted to chars, but a tight budget would starve the summarizer of material — never give it less than this.

api_surface: ApiSurfaceConfig

[context.api_surface] — the workspace-API-surface knowledge_base technique + its pluggable language packs (#669).

input_ceiling_pct: u32

Percent of a request’s num_ctx window usable as INPUT before the pre-send budget gate trims (the rest is reply headroom). The historical hardcoded value was 80 (reserve 20% for the reply). Large-window models (e.g. Opus) can safely run this higher — raising it lets more context ride each turn before trimming kicks in. Clamped to 1..=99; anything outside falls back to 80. See num_ctx_input_ceiling (#282).

low_budget_pct: usize

Percent-of-ceiling below which the loop emits the low-remaining-budget nudge to the model. Historical hardcoded value was 15. Raise it to be warned earlier, lower it to suppress the nudge on roomy models. Clamped to 0..=100; 0 disables the nudge. See agentic::budget (#559).

Trait Implementations

impl Clone for ContextConfig

fn clone(&self) -> ContextConfig

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for ContextConfig

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

Formats the value using the given formatter.

impl Default for ContextConfig

fn default() -> Self

Returns the “default value” for a type.

impl<'de> Deserialize<'de> for ContextConfig

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

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Eq for ContextConfig

impl PartialEq for ContextConfig

fn eq(&self, other: &ContextConfig) -> bool

Equality operator ==.

fn ne(&self, other: &Rhs) -> bool

Inequality operator !=.

impl Serialize for ContextConfig

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

where __S: Serializer,

Serialize this value into the given Serde serializer.

impl StructuralPartialEq for ContextConfig