feat(temporal): refuse disjoint predictions as observed fact

ARCHITECTURE.md

@@ -61,10 +61,10 @@ boundaries above remain the target modular MSA architecture.
 | `tepp_simulation` | known-truth temporal/event data generation |
 | `validation_core` | RMSE, bias, coverage, graph, and Monte Carlo metrics |
 | `tepp_api` | versioned DTO, schema, and export contracts |
+| `prediction_contradiction` | Allen promotion gate: `before`/`after` stay contradictory; `meets`/`met_by` stay unsupported |
 
-No crate exposes placeholder production behavior in Task 1. This prevents an
-empty façade from becoming a de facto public API before its invariants and tests
-exist.
+Foundation crates expose only tested contracts. Empty façades are not public
+APIs.
 
 ## Immutable evidence boundary
 

CHANGELOG.md

@@ -6,6 +6,7 @@ All notable changes to TEPP are documented here. The format follows Keep a Chang
 
 ### Added
 
+- `prediction_contradiction` promotion gate: `temporal_core` Allen classification refuses `before`/`after` as contradiction and `meets`/`met_by` as unsupported adjacency; evidence available after the knowledge cutoff is ineligible. Label agreement is not RMSE recovery (ADR 0002, ADR 0016).
 - `tepp_api` adaptive orchestration router (ADR 0010): versioned `direct`/`verify`/`committee`/`conductor`/`abstain` selection from CPU `f64` risk, ambiguity, evidence, and token-budget inputs; recorded stages, recursion, decomposition, access lists, and role-specific reasoning effort; fail-closed document-controlled policy/access/credentials; LLM plans remain proposals under deterministic statistical authority; comparable-budget ablation requires a direct baseline; credential-free contextual-orchestrator binding. Live NIM HTTP remains accepted-target.
 - `tepp_api` purpose-bound provider-payload minimization: time-bounded `PurposeGrant` evaluation, fail-closed expired/not-yet-valid/inverted/cross-tenant/impossible-calendar denial, semantic UTC calendar validation, refusal to copy identity mappings into model-provider payloads or ordinary logs, preservation of opaque analytical identifiers and membership roles (no blanket PII mask), a separately authorized scientific re-identification path, and an internally bound FIPS 180-4 SHA-256 audit digest appended through `ReidentificationAuditSink` before disclosure.
 - `persistence_postgres` backup/restore integrity: restored snapshots stay unusable until tenant, canonical `SHA-256`, knowledge-cutoff eligibility, temporal window order, and append-only triggers revalidate; SQL probes raise `restore integrity failed` (ADR 0013).

Cargo.toml

@@ -11,6 +11,7 @@ members = [
     "crates/tepp_simulation",
     "crates/validation_core",
     "crates/tepp_api",
+    "crates/prediction_contradiction",
 ]
 default-members = [
     "crates/evidence_core",
@@ -23,6 +24,7 @@ default-members = [
     "crates/tepp_simulation",
     "crates/validation_core",
     "crates/tepp_api",
+    "crates/prediction_contradiction",
 ]
 
 [workspace.package]

README.md

@@ -6,10 +6,11 @@ implemented in Rust.
 
 ## Current implementation state
 
-This branch establishes the Task 1 Rust workspace and quality-gate foundation.
-The ten bounded crates compile independently but intentionally expose no
-placeholder production APIs. Domain behavior begins in Task 2 with immutable
-evidence identifiers and source records.
+This branch keeps the Rust workspace quality foundation and the bounded
+foundation crates. Domain crates expose only tested contracts: immutable
+evidence, six-clock temporal values, event mentions/instances, relations,
+membership, persistence, splits, simulation, validation, API DTOs, and the
+predicted-versus-observed promotion gate.
 
 ```text
 crates/evidence_core
@@ -22,6 +23,7 @@ crates/corpus_split
 crates/tepp_simulation
 crates/validation_core
 crates/tepp_api
+crates/prediction_contradiction
 ```
 
 ## Local verification

crates/prediction_contradiction/Cargo.toml

@@ -0,0 +1,20 @@
+[package]
+name = "prediction_contradiction"
+description = "Predicted intervals that contradict observations cannot become fact."
+version.workspace = true
+edition.workspace = true
+rust-version.workspace = true
+license.workspace = true
+authors.workspace = true
+repository.workspace = true
+homepage.workspace = true
+readme.workspace = true
+keywords.workspace = true
+categories.workspace = true
+publish = false
+
+[dependencies]
+temporal_core = { path = "../temporal_core", version = "0.1.0" }
+
+[lints]
+workspace = true

crates/prediction_contradiction/src/error.rs

@@ -0,0 +1,82 @@
+//! Fail-closed prediction-contradiction errors.
+
+use std::fmt;
+
+/// A fail-closed prediction-contradiction error.
+#[derive(Clone, Copy, Debug, Eq, PartialEq)]
+#[non_exhaustive]
+pub enum PredictionContradictionError {
+    /// Predicted and observed event-time intervals are Allen `before` or `after`.
+    PredictionContradictsObservation,
+    /// Predicted and observed intervals meet but do not overlap in their interiors.
+    PredictionLacksOverlappingSupport,
+    /// Observed evidence covers only part of the predicted interval.
+    PredictionLacksFullSupport,
+    /// Observed evidence became available after the analysis knowledge cutoff.
+    EvidenceAfterCutoff,
+    /// An interval is not a closed proper Allen input.
+    InvalidIntervalPayload,
+    /// An agreement-rate comparison used empty or length-mismatched slices.
+    AgreementSliceMismatch,
+}
+
+impl fmt::Display for PredictionContradictionError {
+    fn fmt(&self, formatter: &mut fmt::Formatter<'_>) -> fmt::Result {
+        let message = match self {
+            Self::PredictionContradictsObservation => {
+                "predicted interval contradicts observed evidence"
+            }
+            Self::PredictionLacksOverlappingSupport => {
+                "predicted interval is adjacent to observation without overlapping support"
+            }
+            Self::PredictionLacksFullSupport => {
+                "predicted interval is not fully supported by observed evidence"
+            }
+            Self::EvidenceAfterCutoff => {
+                "observed evidence is available after the knowledge cutoff"
+            }
+            Self::InvalidIntervalPayload => "invalid prediction-contradiction payload",
+            Self::AgreementSliceMismatch => "agreement slices are empty or length-mismatched",
+        };
+        formatter.write_str(message)
+    }
+}
+
+impl std::error::Error for PredictionContradictionError {}
+
+#[cfg(test)]
+mod tests {
+    use super::PredictionContradictionError;
+
+    #[test]
+    fn error_messages_are_stable() {
+        for (error, message) in [
+            (
+                PredictionContradictionError::PredictionContradictsObservation,
+                "predicted interval contradicts observed evidence",
+            ),
+            (
+                PredictionContradictionError::PredictionLacksOverlappingSupport,
+                "predicted interval is adjacent to observation without overlapping support",
+            ),
+            (
+                PredictionContradictionError::PredictionLacksFullSupport,
+                "predicted interval is not fully supported by observed evidence",
+            ),
+            (
+                PredictionContradictionError::EvidenceAfterCutoff,
+                "observed evidence is available after the knowledge cutoff",
+            ),
+            (
+                PredictionContradictionError::InvalidIntervalPayload,
+                "invalid prediction-contradiction payload",
+            ),
+            (
+                PredictionContradictionError::AgreementSliceMismatch,
+                "agreement slices are empty or length-mismatched",
+            ),
+        ] {
+            assert_eq!(error.to_string(), message);
+        }
+    }
+}