Enum ConfidenceTier 

pub enum ConfidenceTier {
    Confirmed,
    Likely,
    Ambiguous,
}

v0.7.0 Gap 4 (issue #887) — derived enum partitioning the confidence real into operator-meaningful buckets so callers (especially read-side reviewers) can filter by tier instead of re-deriving thresholds at every site.

Thresholds are stable and load-bearing — operators have wired dashboards / human-review queues against them and a change here is a wire-level break. Bumping a threshold is therefore a schema-bump-class decision, NOT a code-tuning decision.

• ConfidenceTier::Confirmed — >= 0.95. High-confidence substrate-curated atoms, typically calibrated by the Form 5 pipeline or asserted by a trusted upstream.
• ConfidenceTier::Likely — 0.7 ..= 0.949…. Default caller-provided observations sit here.
• ConfidenceTier::Ambiguous — < 0.7. The human-review queue: the caller themselves flagged uncertainty (or the decay updater walked the value down). Operators commonly filter their review tool against this tier.

Surfaced to MCP callers via the confidence_calibration.tier_thresholds block on memory_capabilities (Gap 4 read-path closeout).

Disambiguation (issue #970)

The codebase has three enums whose names end in Tier. ConfidenceTier (this enum) is the confidence-value bucket; it is unrelated to:

• Tier — memory-lifecycle TTL bucket (Short/Mid/Long).
• crate::config::FeatureTier — host capability tier (Keyword/Semantic/Smart/Autonomous).

They do not share variants, wire strings, or call sites. See docs/internal/enum-proliferation-audit-970.md.

Variants

Confirmed

Likely

Ambiguous

Implementations

impl ConfidenceTier

pub const CONFIRMED_MIN: f64 = 0.95

Inclusive lower bound for ConfidenceTier::Confirmed. Above this is a high-confidence observation / calibration result.

pub const LIKELY_MIN: f64 = 0.7

Inclusive lower bound for ConfidenceTier::Likely. Below this is the human-review tier (ConfidenceTier::Ambiguous).

pub fn from_confidence(c: f64) -> Self

Bucket a raw confidence value. NaN is conservatively mapped to ConfidenceTier::Ambiguous so a corrupt input lands in the human-review queue rather than masquerading as confirmed.

pub fn as_str(&self) -> &'static str

Wire string for this tier. Matches the serde rename_all = "snake_case" derive above so the JSON and the unstructured helper agree.

pub fn parse(s: &str) -> Option<Self>

Parse a wire string back into the enum. Returns None on unrecognised input so callers can decide whether to error or fall through to “no filter”.

Trait Implementations

impl Clone for ConfidenceTier

fn clone(&self) -> ConfidenceTier

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Copy for ConfidenceTier

impl Debug for ConfidenceTier

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

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for ConfidenceTier

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

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Eq for ConfidenceTier

impl Hash for ConfidenceTier

fn hash<__H: Hasher>(&self, state: &mut __H)

Feeds this value into the given Hasher.

fn hash_slice<H>(data: &[Self], state: &mut H)

where H: Hasher, Self: Sized,

Feeds a slice of this type into the given Hasher.

impl PartialEq for ConfidenceTier

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

Tests for self and other values to be equal, and is used by ==.

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

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl Serialize for ConfidenceTier

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 ConfidenceTier