From e6dc87ea14b443e5a4b457cb0c1e0dae6e120b9c Mon Sep 17 00:00:00 2001 From: seonghobae Date: Tue, 1 Sep 2026 13:32:31 +0000 Subject: [PATCH 1/3] feat(analysis): bind episode-membership refusals to an analysis-run profile GAP-004 leftover / ADR 0072. Bind existing EventWindow and refuse_membership_outside_episode to cutoff-safe episode_membership_v1 (tepp.episode_membership.v1). Mixed contained/escaped census, AvailableTime cutoff exclusion, and MAX_EVIDENCE_UNITS fail closed. identity_recovery_rate stays library-side. Distinct from relation-absence, outcome-order, membership-target, location-membership, and copy-identity. --- CHANGELOG.md | 2 + Cargo.lock | 1 + DOCUMENTATION.md | 1 + crates/analysis_engine/Cargo.toml | 1 + .../src/episode_membership_artifact.rs | 410 ++++++++++++++++++ crates/analysis_engine/src/lib.rs | 15 + .../episode_membership_execution_contract.rs | 252 +++++++++++ docs/TRACEABILITY.md | 1 + .../0072-episode-membership-analysis-run.md | 97 +++++ docs/adr/README.md | 2 + .../episode-membership-analysis-run.md | 18 + 11 files changed, 800 insertions(+) create mode 100644 crates/analysis_engine/src/episode_membership_artifact.rs create mode 100644 crates/analysis_engine/tests/episode_membership_execution_contract.rs create mode 100644 docs/adr/0072-episode-membership-analysis-run.md create mode 100644 docs/doctoring/episode-membership-analysis-run.md diff --git a/CHANGELOG.md b/CHANGELOG.md index 062a69412..2d35931ad 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -38,6 +38,8 @@ All notable changes to TEPP are documented here. The format follows Keep a Chang ## [Unreleased] +- **Episode-membership analysis-run profile**: `analysis_engine` binds existing `episode_membership::EventWindow` and `refuse_membership_outside_episode` to cutoff-safe `episode_membership_v1` (`tepp.episode_membership.v1`) with inference status `membership_window_cannot_escape_episode_interval`. `identity_recovery_rate` stays library-side. Not relation-absence, not outcome-order, not membership-target, not location-membership, not membership-posterior ICC, not copied-text, not copy-identity, not citation-edge, not subevent containment, not GPU, not MCMC, and not topic birth/split/merge. + - `event_core` adds bounded Allen interval-consistency classification, atomic path-consistency closure, contradiction/resource refusals, and an explicit dependency-error fallback without claiming unrestricted global satisfiability. - `psychometric_core` recovers the Driver, Oud, and Voelkle (2017, Table 2, p. 12 `MANIFESTTRAITVAR`; §7.1, p. 19; p. 16 `MANIFESTTRAITVARstd`; footnote 4; 2017-era ctsem `summary.ctsemFit.R`; JSS PDF re-opened 2026-08-27T14:20Z from https://www.jstatsoft.org/index.php/jss/article/download/v077i05/1104) scalar standardised manifest-trait variance on current main after `0ce16e8` dropped the pre-consolidation code while research notes already named the map (register items 83–84). Table 2 names `MANIFESTTRAITVAR` `Ψ_τ` the additional time-invariant variance-covariance on the measurement level and sets it `NULL` when there is no manifest trait. Equation 5 writes `Γ ~ N(τ, Ψ)` and names that covariance the manifest traits. Section 7.1 names manifest traits stable individual differences in indicator levels, distinct from process-level `TRAITVAR` `φ_ξ`. Page 16 prints standardised matrices with the suffix `std` when appropriate. The printed example on p. 16 is `discreteDRIFTstd`, not `MANIFESTTRAITVARstd`. Footnote 4 standardises using only the relevant variance, not the total. The relevant variance for that named indicator-level correlation is `MANIFESTTRAITVAR`, not process-level `TRAITVAR` and not residual `MANIFESTVAR` `θ`. The 2017-era source forms `MANIFESTTRAITVARstd` only when `MANIFESTTRAITVAR != 0`, as `solve(sqrt(diag(MANIFESTTRAITVAR) + ridging)) %&% MANIFESTTRAITVAR` when `verbose = TRUE`. OpenMx `%&%` is `t(A) %*% B %*% A`. Unlike `TRAITVARstd`, that formation adds `diag(c(ridging), n.manifest)`. The default `ridging = FALSE` adds 0, not `0.0001`; that ridge is a numerical hack and is not this exact map. The scalar correlation is `ψ / ψ = 1` after strictly positive `MANIFESTTRAITVAR`. Form strictly positive `ψ` first, then `1 / √ψ`, then `(1 / √ψ) ψ (1 / √ψ)`. Unstandardised `MANIFESTTRAITVAR` is defined for a zero trait; standardised `MANIFESTTRAITVAR` is not. Zero `MANIFESTTRAITVAR` skips forming `MANIFESTTRAITVARstd` in the 2017-era source and fails closed here. Indicator-level trait variance is an event-time structural quantity, so a non-event clock fails closed. `MANIFESTTRAITVAR` does not require stable `a < 0`. Distinct positive `ψ` recover the same 1. `trait / trait = 1` is `TRAITVARstd` and recovers the same number and remains a distinct named quantity. `θ` is `MANIFESTVAR` and is measurement error, not this correlation. Meredith (1993) remains unread (web search 2026-08-27T14:20Z: Springer/Cambridge Core paywalled; Unpaywall historically `is_oa: false`; Springer `content/pdf` is an HTML stub). Mislevy (1991, *Psychometrika, 56*, 177–196) remains unread on the same terms (DOI `10.1007/bf02294457`). Still not a Kalman filter, not a matrix `expm`, not ESEM estimation, not DSEM, and not ctsem estimation. diff --git a/Cargo.lock b/Cargo.lock index 454a7d612..71c6ff138 100644 --- a/Cargo.lock +++ b/Cargo.lock @@ -72,6 +72,7 @@ name = "analysis_engine" version = "0.2.0" dependencies = [ "corpus_split", + "episode_membership", "event_core", "membership_core", "relation_graph", diff --git a/DOCUMENTATION.md b/DOCUMENTATION.md index 6fa4b9683..d72a755f8 100644 --- a/DOCUMENTATION.md +++ b/DOCUMENTATION.md @@ -71,6 +71,7 @@ TEPP's approved PRD v0.4 and implementation plan are the primary product baselin | Hourly NIM OpenCode doctoring | [`docs/doctoring/hourly-nim-opencode-development.md`](docs/doctoring/hourly-nim-opencode-development.md) | | Analysis engine v1 doctoring | [`docs/doctoring/analysis-engine-v1.md`](docs/doctoring/analysis-engine-v1.md) | | Analysis engine gap-closure doctoring | [`docs/doctoring/analysis-engine-gap-closure.md`](docs/doctoring/analysis-engine-gap-closure.md) | +| Episode-membership analysis-run doctoring | [`docs/doctoring/episode-membership-analysis-run.md`](docs/doctoring/episode-membership-analysis-run.md) | | Corpus-split leakage-audit wire doctoring | [`docs/research/corpus-split-manifest-wire.md`](docs/research/corpus-split-manifest-wire.md) | | Unicode canonical-identity doctoring | [`docs/research/unicode-canonical-identity.md`](docs/research/unicode-canonical-identity.md) | | Change history | [`CHANGELOG.md`](CHANGELOG.md) | diff --git a/crates/analysis_engine/Cargo.toml b/crates/analysis_engine/Cargo.toml index 7322212b2..e5f500eba 100644 --- a/crates/analysis_engine/Cargo.toml +++ b/crates/analysis_engine/Cargo.toml @@ -15,6 +15,7 @@ publish = false [dependencies] event_core = { path = "../event_core", version = "0.2.0" } +episode_membership = { path = "../episode_membership", version = "0.2.0" } serde = { workspace = true } serde_json = { workspace = true } sha2 = { workspace = true } diff --git a/crates/analysis_engine/src/episode_membership_artifact.rs b/crates/analysis_engine/src/episode_membership_artifact.rs new file mode 100644 index 000000000..b727c9e22 --- /dev/null +++ b/crates/analysis_engine/src/episode_membership_artifact.rs @@ -0,0 +1,410 @@ +//! Digest-bound episode-membership refusals as an analysis-run profile. + +use episode_membership::{EpisodeMembershipError, EventWindow, refuse_membership_outside_episode}; +use serde::{Deserialize, Serialize}; +use sha2::{Digest, Sha256}; +use temporal_core::{AvailableTime, KnowledgeCutoff}; +use tepp_api::{ + AnalysisResultSummary, AnalysisRunAccepted, AnalysisRunRequest, AnalysisRunTerminalResult, +}; + +use crate::{ + AnalysisEngineError, MAX_EVIDENCE_UNITS, format_digest, require_receipt_identity, + valid_identifier, +}; + +/// Versioned schema for a completed episode-membership artifact. +pub const EPISODE_MEMBERSHIP_ARTIFACT_SCHEMA_VERSION: &str = "tepp.episode_membership.v1"; +/// Model contract required by the episode-membership execution path. +pub const EPISODE_MEMBERSHIP_MODEL_CONTRACT_VERSION: &str = "episode_membership_v1"; +/// Analysis-run output profile required for an episode-membership artifact. +pub const EPISODE_MEMBERSHIP_OUTPUT_PROFILE: &str = "episode_membership_v1"; +/// Maximum canonical artifact JSON size. +pub const EPISODE_MEMBERSHIP_ARTIFACT_BYTE_LIMIT: usize = 256 * 1024; +const EPISODE_MEMBERSHIP_INFERENCE_STATUS: &str = + "membership_window_cannot_escape_episode_interval"; + +/// One cutoff-admitted membership assignment against an episode window. +#[derive(Clone, Debug, Eq, PartialEq)] +pub struct EpisodeMembershipAssignment { + assignment_id: String, + membership: EventWindow, + episode: EventWindow, + available_time: AvailableTime, +} + +impl EpisodeMembershipAssignment { + /// Construct a bounded episode-membership assignment. + /// + /// # Errors + /// + /// Returns [`AnalysisEngineError::InvalidEvidence`] when the assignment + /// identity is empty or oversized. + pub fn new( + assignment_id: impl Into, + membership: EventWindow, + episode: EventWindow, + available_time: AvailableTime, + ) -> Result { + let assignment_id = assignment_id.into(); + if !valid_identifier(&assignment_id) { + return Err(AnalysisEngineError::InvalidEvidence); + } + Ok(Self { + assignment_id, + membership, + episode, + available_time, + }) + } + + /// Return the opaque assignment identity. + #[must_use] + pub fn assignment_id(&self) -> &str { + &self.assignment_id + } + + /// Return the membership event-time window. + #[must_use] + pub const fn membership(&self) -> EventWindow { + self.membership + } + + /// Return the episode event-time window. + #[must_use] + pub const fn episode(&self) -> EventWindow { + self.episode + } + + /// Return the availability time used for cutoff eligibility. + #[must_use] + pub const fn available_time(&self) -> AvailableTime { + self.available_time + } +} + +/// Completed, bounded episode-membership census for analysis-run clients. +#[derive(Clone, Debug, Deserialize, PartialEq, Serialize)] +#[serde(deny_unknown_fields)] +pub struct EpisodeMembershipArtifact { + /// Exact versioned schema identity. + pub schema_version: String, + /// Opaque accepted-run identity. + pub run_id: String, + /// Immutable source snapshot identity. + pub snapshot_id: String, + /// Historical evidence cutoff used to admit assignments. + pub knowledge_cutoff: String, + /// Number of assignments admitted at the cutoff. + pub assignment_count: u64, + /// Assignments contained in their episode window. + pub contained_count: u64, + /// Assignments that escaped their episode window. + pub escaped_count: u64, + /// Escaped assignments refused as membership-outside-episode. + pub refused_as_escape_count: u64, + /// Fixed claim boundary for consumer copy. + pub inference_status: String, +} + +impl EpisodeMembershipArtifact { + /// Parse and fully validate a bounded artifact JSON payload. + /// + /// # Errors + /// + /// Returns [`AnalysisEngineError::InvalidEpisodeMembershipArtifact`] when + /// the schema, identifiers, counts, or claim boundary fail. + pub fn from_json(payload: &str) -> Result { + if payload.len() > EPISODE_MEMBERSHIP_ARTIFACT_BYTE_LIMIT { + return Err(AnalysisEngineError::LimitExceeded); + } + let artifact: Self = serde_json::from_str(payload) + .map_err(|_| AnalysisEngineError::InvalidEpisodeMembershipArtifact)?; + artifact.validate()?; + Ok(artifact) + } + + /// Serialize canonical validated artifact JSON. + /// + /// # Errors + /// + /// Returns a typed validation, serialization, or size failure. + pub fn to_json(&self) -> Result { + self.validate()?; + let payload = + serde_json::to_string(self).map_err(|_| AnalysisEngineError::SerializationFailure)?; + if payload.len() > EPISODE_MEMBERSHIP_ARTIFACT_BYTE_LIMIT { + return Err(AnalysisEngineError::LimitExceeded); + } + Ok(payload) + } + + /// Return the lowercase SHA-256 digest of canonical artifact JSON. + /// + /// # Errors + /// + /// Returns a typed validation or serialization failure. + pub fn sha256(&self) -> Result { + self.to_json() + .map(|json| format_digest(Sha256::digest(json.into_bytes()))) + } + + fn validate(&self) -> Result<(), AnalysisEngineError> { + let status_sum = self.contained_count.checked_add(self.escaped_count); + if self.schema_version != EPISODE_MEMBERSHIP_ARTIFACT_SCHEMA_VERSION + || !valid_identifier(&self.run_id) + || !valid_identifier(&self.snapshot_id) + || KnowledgeCutoff::parse_rfc3339(&self.knowledge_cutoff).is_err() + || self.assignment_count < 2 + || self.contained_count == 0 + || self.escaped_count == 0 + || status_sum != Some(self.assignment_count) + || self.refused_as_escape_count != self.escaped_count + || self.inference_status != EPISODE_MEMBERSHIP_INFERENCE_STATUS + { + return Err(AnalysisEngineError::InvalidEpisodeMembershipArtifact); + } + Ok(()) + } +} + +/// One completed episode-membership artifact and its terminal result. +#[derive(Clone, Debug, PartialEq)] +pub struct EpisodeMembershipExecution { + /// Digest-bound completed episode-membership census. + pub artifact: EpisodeMembershipArtifact, + /// Terminal result carrying the artifact identity, digest, and schema. + pub terminal_result: AnalysisRunTerminalResult, +} + +/// Execute cutoff-safe episode-membership refusals as one analysis-run profile. +/// +/// The executor invokes [`refuse_membership_outside_episode`] already on +/// protected main. Contained windows stay membership. Escaped windows stay +/// refusals, never subevent-versus-parent containment. It does not emit +/// `identity_recovery_rate`, a `scientific_acceptance` inspect metric, GPU +/// kernels, MCMC, or topic birth/split/merge events. +/// +/// # Errors +/// +/// Returns a request/receipt/snapshot/cutoff/profile error, empty or +/// single-class corpus, inverted or escaped membership treated as success, +/// duplicate assignment identity, oversized corpus, or invalid artifact +/// error. +pub fn execute_episode_membership_run( + request: &AnalysisRunRequest, + accepted: &AnalysisRunAccepted, + snapshot_id: &str, + knowledge_cutoff: KnowledgeCutoff, + assignments: &[EpisodeMembershipAssignment], + completed_at: impl Into, +) -> Result { + request.to_json()?; + accepted.to_json()?; + require_receipt_identity(request, accepted)?; + if request.snapshot_id != snapshot_id { + return Err(AnalysisEngineError::SnapshotMismatch); + } + if request.knowledge_cutoff != knowledge_cutoff.to_rfc3339() + || request.model_contract_version != EPISODE_MEMBERSHIP_MODEL_CONTRACT_VERSION + || request.output_profile != EPISODE_MEMBERSHIP_OUTPUT_PROFILE + { + return Err(AnalysisEngineError::InvalidEvidence); + } + if assignments.len() > MAX_EVIDENCE_UNITS { + return Err(AnalysisEngineError::LimitExceeded); + } + + let (contained_count, escaped_count, refused_as_escape_count) = + census_assignments(assignments, knowledge_cutoff)?; + let assignment_count = contained_count + .checked_add(escaped_count) + .ok_or(AnalysisEngineError::ArithmeticOverflow)?; + if assignment_count < 2 + || contained_count == 0 + || escaped_count == 0 + || refused_as_escape_count != escaped_count + { + return Err(AnalysisEngineError::InvalidEvidence); + } + + let artifact = EpisodeMembershipArtifact { + schema_version: EPISODE_MEMBERSHIP_ARTIFACT_SCHEMA_VERSION.into(), + run_id: accepted.run_id.clone(), + snapshot_id: snapshot_id.to_owned(), + knowledge_cutoff: knowledge_cutoff.to_rfc3339(), + assignment_count, + contained_count, + escaped_count, + refused_as_escape_count, + inference_status: EPISODE_MEMBERSHIP_INFERENCE_STATUS.into(), + }; + let digest = artifact.sha256()?; + let summary = AnalysisResultSummary::new( + "episode_membership", + assignment_count, + 4, + EPISODE_MEMBERSHIP_INFERENCE_STATUS, + )?; + let terminal_result = AnalysisRunTerminalResult::succeeded( + request, + accepted, + format!("episode_membership_artifact_{}", &digest[..16]), + digest, + EPISODE_MEMBERSHIP_ARTIFACT_SCHEMA_VERSION, + completed_at, + summary, + )?; + Ok(EpisodeMembershipExecution { + artifact, + terminal_result, + }) +} + +fn census_assignments( + assignments: &[EpisodeMembershipAssignment], + knowledge_cutoff: KnowledgeCutoff, +) -> Result<(u64, u64, u64), AnalysisEngineError> { + let mut seen = std::collections::BTreeSet::new(); + let mut contained_count = 0_u64; + let mut escaped_count = 0_u64; + let mut refused_as_escape_count = 0_u64; + for assignment in assignments { + if !seen.insert(assignment.assignment_id()) { + return Err(AnalysisEngineError::DuplicateEvidence); + } + if assignment.available_time().instant() > knowledge_cutoff.instant() { + continue; + } + match refuse_membership_outside_episode(assignment.membership(), assignment.episode()) { + Ok(()) => { + contained_count = increment(contained_count)?; + } + Err(EpisodeMembershipError::MembershipEscapesEpisode) => { + refused_as_escape_count = increment(refused_as_escape_count)?; + escaped_count = increment(escaped_count)?; + } + Err( + EpisodeMembershipError::InvertedEventWindow + | EpisodeMembershipError::InvalidEpisodePayload + | _, + ) => return Err(AnalysisEngineError::InvalidEvidence), + } + } + Ok((contained_count, escaped_count, refused_as_escape_count)) +} + +fn increment(count: u64) -> Result { + count + .checked_add(1) + .ok_or(AnalysisEngineError::ArithmeticOverflow) +} + +#[cfg(test)] +mod tests { + use super::{ + EPISODE_MEMBERSHIP_ARTIFACT_BYTE_LIMIT, EPISODE_MEMBERSHIP_ARTIFACT_SCHEMA_VERSION, + EPISODE_MEMBERSHIP_INFERENCE_STATUS, EpisodeMembershipArtifact, + }; + use crate::AnalysisEngineError; + + fn artifact() -> EpisodeMembershipArtifact { + EpisodeMembershipArtifact { + schema_version: EPISODE_MEMBERSHIP_ARTIFACT_SCHEMA_VERSION.into(), + run_id: "run-1".into(), + snapshot_id: "snapshot-1".into(), + knowledge_cutoff: "2026-08-01T00:00:00Z".into(), + assignment_count: 2, + contained_count: 1, + escaped_count: 1, + refused_as_escape_count: 1, + inference_status: EPISODE_MEMBERSHIP_INFERENCE_STATUS.into(), + } + } + + fn assert_invalid(artifact: &EpisodeMembershipArtifact) { + assert_eq!( + artifact.to_json(), + Err(AnalysisEngineError::InvalidEpisodeMembershipArtifact) + ); + } + + #[test] + fn artifact_round_trip_and_size_bounds_fail_closed() { + let artifact = artifact(); + let payload = artifact.to_json().expect("json"); + assert_eq!( + EpisodeMembershipArtifact::from_json(&payload), + Ok(artifact.clone()) + ); + assert_eq!(artifact.sha256().expect("digest").len(), 64); + assert_eq!( + EpisodeMembershipArtifact::from_json("{}"), + Err(AnalysisEngineError::InvalidEpisodeMembershipArtifact) + ); + assert_eq!( + EpisodeMembershipArtifact::from_json( + &"x".repeat(EPISODE_MEMBERSHIP_ARTIFACT_BYTE_LIMIT + 1) + ), + Err(AnalysisEngineError::LimitExceeded) + ); + } + + #[test] + fn artifact_metadata_tampering_fails_closed() { + let artifact = artifact(); + let invalid_artifacts = [ + { + let mut value = artifact.clone(); + value.schema_version.clear(); + value + }, + { + let mut value = artifact.clone(); + value.run_id.clear(); + value + }, + { + let mut value = artifact.clone(); + value.snapshot_id.clear(); + value + }, + { + let mut value = artifact.clone(); + value.knowledge_cutoff = "invalid".into(); + value + }, + { + let mut value = artifact.clone(); + value.assignment_count = 1; + value + }, + { + let mut value = artifact.clone(); + value.contained_count = 0; + value.assignment_count = 1; + value + }, + { + let mut value = artifact.clone(); + value.escaped_count = 0; + value.refused_as_escape_count = 0; + value.assignment_count = 1; + value + }, + { + let mut value = artifact.clone(); + value.refused_as_escape_count = 0; + value + }, + { + let mut value = artifact.clone(); + value.inference_status.clear(); + value + }, + ]; + for invalid in invalid_artifacts { + assert_invalid(&invalid); + } + } +} diff --git a/crates/analysis_engine/src/lib.rs b/crates/analysis_engine/src/lib.rs index 72bd5854c..19a9942d8 100644 --- a/crates/analysis_engine/src/lib.rs +++ b/crates/analysis_engine/src/lib.rs @@ -11,6 +11,7 @@ //! contracts and preserves their artifact meaning. mod case_deletion_refit; +mod episode_membership_artifact; mod lineage_criterion; mod topic_context_posterior; mod topic_lineage_artifact; @@ -41,6 +42,13 @@ pub use case_deletion_refit::ExhaustiveCaseDeletionError; pub use case_deletion_refit::ExhaustiveCaseDeletionFits; /// Fit the full corpus and every actual one-document deletion. pub use case_deletion_refit::fit_exhaustive_case_deletion; +/// Episode-membership artifact and execution contracts from this engine. +pub use episode_membership_artifact::{ + EPISODE_MEMBERSHIP_ARTIFACT_BYTE_LIMIT, EPISODE_MEMBERSHIP_ARTIFACT_SCHEMA_VERSION, + EPISODE_MEMBERSHIP_MODEL_CONTRACT_VERSION, EPISODE_MEMBERSHIP_OUTPUT_PROFILE, + EpisodeMembershipArtifact, EpisodeMembershipAssignment, EpisodeMembershipExecution, + execute_episode_membership_run, +}; /// Rust-owned independent TDT link-criterion posterior fitting contracts. pub use lineage_criterion::{ LineageCriterionFit, LineageCriterionFitError, LineageCriterionObservation, @@ -248,6 +256,8 @@ pub enum AnalysisEngineError { TopicMeasurement(TopicMeasurementError), /// A topic-lineage artifact violated its bounded schema or count invariants. InvalidTopicLineageArtifact, + /// An episode-membership artifact violated its bounded schema or count invariants. + InvalidEpisodeMembershipArtifact, } impl fmt::Display for AnalysisEngineError { @@ -262,6 +272,7 @@ impl fmt::Display for AnalysisEngineError { Self::LimitExceeded => "analysis corpus exceeded its execution bound", Self::TopicMeasurement(error) => return error.fmt(formatter), Self::InvalidTopicLineageArtifact => "invalid topic lineage artifact", + Self::InvalidEpisodeMembershipArtifact => "invalid episode-membership artifact", }; formatter.write_str(message) } @@ -681,6 +692,10 @@ mod tests { AnalysisEngineError::InvalidTopicLineageArtifact, "invalid topic lineage artifact", ), + ( + AnalysisEngineError::InvalidEpisodeMembershipArtifact, + "invalid episode-membership artifact", + ), ]; for (error, message) in messages { assert_eq!(error.to_string(), message); diff --git a/crates/analysis_engine/tests/episode_membership_execution_contract.rs b/crates/analysis_engine/tests/episode_membership_execution_contract.rs new file mode 100644 index 000000000..093df7c29 --- /dev/null +++ b/crates/analysis_engine/tests/episode_membership_execution_contract.rs @@ -0,0 +1,252 @@ +//! End-to-end contract for cutoff-safe episode-membership refusals. + +use analysis_engine::{ + AnalysisEngineError, EPISODE_MEMBERSHIP_ARTIFACT_SCHEMA_VERSION, + EPISODE_MEMBERSHIP_MODEL_CONTRACT_VERSION, EPISODE_MEMBERSHIP_OUTPUT_PROFILE, + EpisodeMembershipAssignment, MAX_EVIDENCE_UNITS, execute_episode_membership_run, +}; +use episode_membership::EventWindow; +use temporal_core::{AvailableTime, KnowledgeCutoff}; +use tepp_api::{AnalysisRunAccepted, AnalysisRunRequest, AnalysisRunTerminalState}; + +fn cutoff() -> KnowledgeCutoff { + KnowledgeCutoff::parse_rfc3339("2026-08-01T00:00:00Z").expect("cutoff") +} + +fn available(stamp: &str) -> AvailableTime { + AvailableTime::parse_rfc3339(stamp).expect("available") +} + +fn window(start: i64, end: i64) -> EventWindow { + EventWindow::new(start, end).expect("window") +} + +fn request() -> AnalysisRunRequest { + AnalysisRunRequest { + contract_version: 1, + idempotency_key: "episode-membership-idem".into(), + tenant_workspace_id: "tenant-workspace".into(), + snapshot_id: "snapshot-episode-membership".into(), + knowledge_cutoff: "2026-08-01T00:00:00Z".into(), + model_contract_version: EPISODE_MEMBERSHIP_MODEL_CONTRACT_VERSION.into(), + output_profile: EPISODE_MEMBERSHIP_OUTPUT_PROFILE.into(), + } +} + +fn accepted(request: &AnalysisRunRequest) -> AnalysisRunAccepted { + AnalysisRunAccepted::new( + "run-episode-membership", + "accepted", + &request.idempotency_key, + ) + .expect("accepted") +} + +fn assignment( + assignment_id: &str, + membership: EventWindow, + episode: EventWindow, + stamp: &str, +) -> EpisodeMembershipAssignment { + EpisodeMembershipAssignment::new(assignment_id, membership, episode, available(stamp)) + .expect("assignment") +} + +fn mixed_assignments() -> Vec { + let episode = window(10, 20); + vec![ + assignment( + "contained-a", + window(11, 19), + episode, + "2026-07-01T00:00:00Z", + ), + assignment("escaped-b", window(9, 15), episode, "2026-07-02T00:00:00Z"), + ] +} + +fn execute( + request: &AnalysisRunRequest, + assignments: &[EpisodeMembershipAssignment], +) -> Result { + execute_episode_membership_run( + request, + &accepted(request), + "snapshot-episode-membership", + cutoff(), + assignments, + "2026-08-02T00:00:00Z", + ) +} + +#[test] +fn mixed_windows_emit_digest_bound_refusals_without_recovery_metric() { + let request = request(); + let execution = execute(&request, &mixed_assignments()).expect("execution"); + assert_eq!( + execution.artifact.schema_version, + EPISODE_MEMBERSHIP_ARTIFACT_SCHEMA_VERSION + ); + assert_eq!(execution.artifact.assignment_count, 2); + assert_eq!(execution.artifact.contained_count, 1); + assert_eq!(execution.artifact.escaped_count, 1); + assert_eq!(execution.artifact.refused_as_escape_count, 1); + assert_eq!( + execution.artifact.inference_status, + "membership_window_cannot_escape_episode_interval" + ); + let payload = execution.artifact.to_json().expect("json"); + assert!(!payload.contains("identity_recovery_rate")); + assert!(!payload.contains("scientific_acceptance")); + assert!(!payload.contains("subevent")); + assert_eq!( + execution.terminal_result.run_state, + AnalysisRunTerminalState::Succeeded + ); + assert_eq!( + execution.terminal_result.result_sha256.as_deref(), + Some(execution.artifact.sha256().expect("digest").as_str()) + ); + assert_eq!( + execution.terminal_result.result_schema_version.as_deref(), + Some(EPISODE_MEMBERSHIP_ARTIFACT_SCHEMA_VERSION) + ); +} + +#[test] +fn future_available_assignments_are_excluded() { + let request = request(); + let episode = window(10, 20); + let mut with_future = mixed_assignments(); + with_future.push(assignment( + "future-c", + window(11, 12), + episode, + "2026-08-02T00:00:00Z", + )); + let execution = execute(&request, &with_future).expect("cutoff"); + assert_eq!(execution.artifact.assignment_count, 2); + assert_eq!(execution.artifact.contained_count, 1); +} + +#[test] +fn empty_or_single_class_and_duplicate_fail_closed() { + let request = request(); + let episode = window(10, 20); + let stamp = "2026-07-01T00:00:00Z"; + assert_eq!( + execute(&request, &[]), + Err(AnalysisEngineError::InvalidEvidence) + ); + let contained_only = vec![ + assignment("contained-a", window(11, 12), episode, stamp), + assignment("contained-b", window(13, 14), episode, stamp), + ]; + assert_eq!( + execute(&request, &contained_only), + Err(AnalysisEngineError::InvalidEvidence) + ); + let escaped_only = vec![ + assignment("escaped-a", window(1, 5), episode, stamp), + assignment("escaped-b", window(21, 25), episode, stamp), + ]; + assert_eq!( + execute(&request, &escaped_only), + Err(AnalysisEngineError::InvalidEvidence) + ); + let duplicates = vec![ + assignment("same", window(11, 12), episode, stamp), + assignment("same", window(1, 5), episode, stamp), + ]; + assert_eq!( + execute(&request, &duplicates), + Err(AnalysisEngineError::DuplicateEvidence) + ); + assert_eq!( + EpisodeMembershipAssignment::new("", window(11, 12), episode, available(stamp)), + Err(AnalysisEngineError::InvalidEvidence) + ); +} + +#[test] +fn execution_refuses_snapshot_profile_cutoff_mismatch_and_oversize() { + let request = request(); + let assignments = mixed_assignments(); + assert_eq!( + execute_episode_membership_run( + &request, + &accepted(&request), + "other-snapshot", + cutoff(), + &assignments, + "2026-08-02T00:00:00Z", + ), + Err(AnalysisEngineError::SnapshotMismatch) + ); + let mut mismatched = request.clone(); + mismatched.knowledge_cutoff = "2026-07-01T00:00:00Z".into(); + assert_eq!( + execute_episode_membership_run( + &mismatched, + &accepted(&mismatched), + "snapshot-episode-membership", + cutoff(), + &assignments, + "2026-08-02T00:00:00Z", + ), + Err(AnalysisEngineError::InvalidEvidence) + ); + for profile in [ + "trsl_topic_lineage_v1", + "fitted_candidate_k_v1", + "pareto_candidate_k_v1", + "joint_posterior_draws_v1", + "method_effects_v1", + "copy_identity_v1", + "style_source_v1", + "prompt_source_v1", + "modality_source_v1", + "corpus_background_v1", + "citation_edge_v1", + "copied_text_v1", + "lineage_criterion_v1", + "composed_fitted_lineage_v1", + "case_deletion_refit_v1", + "topic_activity_v1", + "location_membership_v1", + "topic_context_posterior_v1", + "membership_posterior_icc_v1", + "membership_target_v1", + "outcome_order_v1", + "relation_absence_v1", + ] { + let mut reused = request.clone(); + reused.output_profile = profile.into(); + assert_eq!( + execute_episode_membership_run( + &reused, + &accepted(&reused), + "snapshot-episode-membership", + cutoff(), + &assignments, + "2026-08-02T00:00:00Z", + ), + Err(AnalysisEngineError::InvalidEvidence) + ); + } + let episode = window(10, 20); + let oversized: Vec = (0..=MAX_EVIDENCE_UNITS) + .map(|index| { + assignment( + &format!("assignment-{index}"), + window(11, 12), + episode, + "2026-07-01T00:00:00Z", + ) + }) + .collect(); + assert_eq!( + execute(&request, &oversized), + Err(AnalysisEngineError::LimitExceeded) + ); +} diff --git a/docs/TRACEABILITY.md b/docs/TRACEABILITY.md index 2b783c2ab..f96be4f53 100644 --- a/docs/TRACEABILITY.md +++ b/docs/TRACEABILITY.md @@ -58,6 +58,7 @@ The full APA 7th standards/literature register remains `docs/research/standards- | versioned service/API contracts and exports | PRD; API contract; ADR 0011/0013 | `tepp_api` analysis-run/export/JSON-LD/GraphML contracts on protected main (PR #21); HTTP service remaining accepted-target | partial | | versioned service/API contracts and exports | PRD; API contract; ADR 0011/0013 | `tepp_api` analysis-run/export/JSON-LD/GraphML contracts on protected main (PR #21); LineageWeave loopback contracts and request-bound terminal result are composed on the active product branch; production TLS remaining | partial | | executable cutoff-safe analysis runs | ADR 0012/0022; temporal research; API terminal-result contract | `analysis_engine` availability cutoff, snapshot binding, multiple-membership aggregation, digest-bound readiness artifact, and `tepp.trsl_topic_lineage.v1` execution through `topic_measurement`; synthetic recovery plus tamper/non-convergence tests and exact coverage on the active product branch | active-PR | +| episode-membership analysis-run profile | ADR 0003/0022/0072; membership windows cannot escape the episode interval | `analysis_engine` `episode_membership_v1` binds `refuse_membership_outside_episode`; digest-bound refusals, not `identity_recovery_rate` inspect metric, not relation-absence, not outcome-order, not membership-target, not location-membership, not subevent containment, not GPU, not MCMC, not topic birth/split/merge; not implemented-main | active-PR | | immutable split/run/reproducibility manifests | ADR 0013; ERD | `tepp_api` reproducibility manifest contract on protected main; `persistence_postgres` append-only SQL insert/lookup for `reproducibility_manifest`, `corpus_split_manifest`, `model_run`, and `model_artifact` (migration `0003`); full physical ERD constraints remaining | partial | | multilingual shared latent semantic space | PRD; ADR 0004; ADR 0020 | `semantic_core` span-grounded units (active-PR); concept dictionary and shared latent estimator remaining | active-PR | | TRSL-TM temporal/relational topic posterior and backend compatibility | ADR 0012; ADR 0004 | `topic_measurement` stable ALR/ILR coordinates and bounded CPU `f64` reference estimator on protected main; `model_selection` fitted candidate-`K` scoring on this PR; calibrated posterior promotion, method effects, persistence, and accelerated backends remaining | partial | diff --git a/docs/adr/0072-episode-membership-analysis-run.md b/docs/adr/0072-episode-membership-analysis-run.md new file mode 100644 index 000000000..398d231e8 --- /dev/null +++ b/docs/adr/0072-episode-membership-analysis-run.md @@ -0,0 +1,97 @@ +# ADR 0072 — Episode-membership refusals as an analysis-run output profile + +**Decision status:** Accepted +**Implementation maturity:** active-PR — composed on this branch; not implemented-main +**Date:** 2026-09-01 +**Supersedes:** None; complements ADR 0003 (episode membership cannot escape the episode event-time interval) and ADR 0022 (cutoff-safe analysis-run execution). Does not reuse ADR 0071 (relation-absence), ADR 0070 (outcome-order), ADR 0069 (membership-target), ADR 0068 (topic-context posterior), ADR 0066 (location-membership), ADR 0065 (copied-text residue), ADR 0064 (provenance-is-not-transition / citation-edge), or ADR 0058 (copy-identity / template-copy). This is membership-window containment, not subevent-versus-parent containment. +**Figma File ID:** N/A — this increment changes a Rust service crate and has no user-interface surface. +**Storybook inventory:** N/A — no reusable web object or interaction changed. + +## Context + +Protected main already refuses a membership window that starts before or +ends after its episode, via `episode_membership::EventWindow` and +`refuse_membership_outside_episode`. Operators still cannot request that +census as a digest-bound analysis-run output. + +Relation-absence (#460 / ADR 0071) binds observation status. +Outcome-order (#458 / ADR 0070) binds IPO event-time order. +Membership-target (#434 / ADR 0069) binds `MembershipTargetKind`. +Location-membership (#430 / ADR 0066) binds geographic/market assignment. + +`identity_recovery_rate` stays library-side. This slice does not put a +`scientific_acceptance` metric on inspect payloads. Subevent parent-window +containment remains `subevent_containment`. + +GPU kernels, MCMC, and topic birth/split/merge remain later GAP-004 work +and are not this slice. + +## Decision + +Add the `episode_membership_v1` analysis-run output profile to +`analysis_engine`. The executor: + +- consumes already-validated `EpisodeMembershipAssignment` rows with + closed `EventWindow` membership/episode bounds and availability time; +- requires the request snapshot and knowledge cutoff to match the offered + input construction; +- excludes assignments whose availability is later than the knowledge + cutoff; +- invokes `refuse_membership_outside_episode` without reimplementing the + containment vocabulary; +- requires a mixed census of at least one contained and one escaped + assignment after cutoff exclusion; +- emits a canonical SHA-256-digested `tepp.episode_membership.v1` artifact + with contained/escaped counts, matching escape-refusal counts, and + inference status `membership_window_cannot_escape_episode_interval`; +- does not emit `identity_recovery_rate`, invent MCMC, select GPU + backends, or emit topic birth/split/merge events. + +## Alternatives considered + +1. Duplicate relation-absence (#460 / ADR 0071) — rejected because that + profile binds observation status, not episode-window containment. +2. Duplicate outcome-order (#458 / ADR 0070) — rejected because that + profile binds IPO event-time order. +3. Duplicate membership-target (#434 / ADR 0069) — rejected because that + profile binds `MembershipTargetKind`. +4. Duplicate location-membership (#430 / ADR 0066) — rejected because + that profile binds `location_membership` LocationKind refusals. +5. Put `identity_recovery_rate` on the operator artifact — rejected + because inspect payloads stay metric-free and + `tepp.scientific_acceptance.v1` never appears. +6. Bind the existing episode-membership refusals to ADR 0022's + analysis-run profile — accepted. + +## Consequences + +Operators can request cutoff-safe episode-membership refusals as a +digest-bound terminal result. The artifact does not claim MCMC, GPU +parity, relation-absence, outcome-order, membership-target, +location-membership, membership-posterior ICC, copied-text, +copy-identity, citation-edge, subevent containment, method-effect +estimation, or topic birth/split/merge. Snapshot / profile / cutoff +mismatch, empty or single-class corpora, duplicate assignment +identities, and oversized corpora fail closed. + +## Verification + +The PR includes Rust unit and integration tests for mixed +contained/escaped corpora, cutoff exclusion, empty/single-class/duplicate +refusal, snapshot / profile / cutoff mismatch, oversize, and artifact +tampering. Run: + +```text +cargo fmt --all -- --check +cargo test -p analysis_engine +cargo clippy -p analysis_engine --all-targets -- -D warnings +python3 scripts/validate_documentation.py +``` + +## Rollback and supersession + +Rollback removes the `episode_membership_v1` profile. No persisted schema +migration is introduced. Supersede only with an ADR that keeps membership +windows inside the episode interval, keeps this distinct from +subevent-versus-parent containment, and keeps `identity_recovery_rate` +off inspect payloads. diff --git a/docs/adr/README.md b/docs/adr/README.md index 1254c8079..15005b7bd 100644 --- a/docs/adr/README.md +++ b/docs/adr/README.md @@ -30,6 +30,7 @@ Read [`ADR_POLICY.md`](ADR_POLICY.md) first. **Decision status and implementatio | [0022](0022-deterministic-analysis-run-execution.md) | Deterministic cutoff-safe analysis-run execution | Accepted | active-PR | Closes the first executable product path from accepted run to digest-bound terminal result without claiming estimator authority. | | [0024](0024-lineage-pair-criterion-and-project-journey-posterior.md) | Independent Event Lineage pair criterion and posterior Project Journey | Proposed | active-PR | Strict artifacts preserve criterion/event-time draws, branches, ties, and CPU/GPU receipts without claiming the scientific estimator is complete. | | [0025](0025-macos-native-rust-mlx-metal-boundary.md) | macOS-native Rust-owned MLX Metal execution | Accepted | accepted-target | Compose authenticates to a native host service; Linux never claims Metal, and actual backend/parity receipts fail closed. | +| [0072](0072-episode-membership-analysis-run.md) | Episode-membership refusals as an analysis-run profile | Accepted | active-PR | Complements ADR 0003/0022; `EventWindow` + `refuse_membership_outside_episode`, not relation-absence, not outcome-order, not membership-target. | | [0023](0023-lineage-criterion-anchor-contract.md) | TEPP-owned Event Lineage criterion anchor | Accepted | active-PR | PR #237 publishes the strict accepted/rejected artifact and identities; estimator execution remains fail-closed future work. | | [0024](0024-independent-topic-importance-anchor.md) | Posterior topic-context producer contract | Accepted | contract-only active-PR | Strict DTO/schema only; the current estimator does not emit it. fast-mlsirm owns case-deletion influence. | | [0001](0001-rust-first-modular-msa.md) | Rust-first numerical core and CPU `f64` reference | Accepted | partial | ADR 0011 owns cross-service/MSA authority; 0001 retains numerical/backend authority. | @@ -138,6 +139,7 @@ Use the narrowest owning ADR when decisions overlap: - **project-history wire-size symmetry:** ADR 0019. - **LineageWeave project-history service boundary:** ADR 0021. - **accepted-run execution and terminal artifact production:** ADR 0022. +- **episode-membership analysis-run profile:** ADR 0072. - **independent lineage criterion and posterior Project Journey:** ADR 0023. - **macOS-native Rust-owned MLX Metal execution:** ADR 0024. diff --git a/docs/doctoring/episode-membership-analysis-run.md b/docs/doctoring/episode-membership-analysis-run.md new file mode 100644 index 000000000..b6b50888a --- /dev/null +++ b/docs/doctoring/episode-membership-analysis-run.md @@ -0,0 +1,18 @@ +# Episode-membership analysis-run composition + +**Active slice:** ADR 0072 / `episode_membership_v1` +**Protected-main status:** not implemented-main + +`episode_membership` already refuses a membership window that starts +before or ends after its episode. This slice binds `EventWindow` and +`refuse_membership_outside_episode` to a cutoff-safe analysis-run +profile so operators can request a digest-bound identity artifact. + +The artifact inference status is +`membership_window_cannot_escape_episode_interval`. +`identity_recovery_rate` stays library-side. This is membership-window +containment, not subevent-versus-parent containment. This is not +relation-absence, not outcome-order, not membership-target, not +location-membership, not membership-posterior ICC, not copied-text, not +copy-identity, not citation-edge, not GPU, not MCMC, and not topic +birth/split/merge. From 765e38c85621e054c01a3b8670ec1cdf5e3f17af Mon Sep 17 00:00:00 2001 From: Seongho Bae Date: Wed, 2 Sep 2026 07:07:42 +0900 Subject: [PATCH 2/3] test(analysis): reject impossible compact membership census --- .../episode_membership_execution_contract.rs | 29 ++++++++++++++++++- 1 file changed, 28 insertions(+), 1 deletion(-) diff --git a/crates/analysis_engine/tests/episode_membership_execution_contract.rs b/crates/analysis_engine/tests/episode_membership_execution_contract.rs index 093df7c29..8041e0ad4 100644 --- a/crates/analysis_engine/tests/episode_membership_execution_contract.rs +++ b/crates/analysis_engine/tests/episode_membership_execution_contract.rs @@ -3,7 +3,8 @@ use analysis_engine::{ AnalysisEngineError, EPISODE_MEMBERSHIP_ARTIFACT_SCHEMA_VERSION, EPISODE_MEMBERSHIP_MODEL_CONTRACT_VERSION, EPISODE_MEMBERSHIP_OUTPUT_PROFILE, - EpisodeMembershipAssignment, MAX_EVIDENCE_UNITS, execute_episode_membership_run, + EpisodeMembershipArtifact, EpisodeMembershipAssignment, MAX_EVIDENCE_UNITS, + execute_episode_membership_run, }; use episode_membership::EventWindow; use temporal_core::{AvailableTime, KnowledgeCutoff}; @@ -113,6 +114,32 @@ fn mixed_windows_emit_digest_bound_refusals_without_recovery_metric() { ); } +#[test] +fn compact_oversized_artifact_counts_fail_closed() { + let assignment_count = MAX_EVIDENCE_UNITS as u64 + 1; + let escaped_count = assignment_count - 1; + let artifact = EpisodeMembershipArtifact { + schema_version: EPISODE_MEMBERSHIP_ARTIFACT_SCHEMA_VERSION.into(), + run_id: "run-compact-oversize".into(), + snapshot_id: "snapshot-compact-oversize".into(), + knowledge_cutoff: "2026-08-01T00:00:00Z".into(), + assignment_count, + contained_count: 1, + escaped_count, + refused_as_escape_count: escaped_count, + inference_status: "membership_window_cannot_escape_episode_interval".into(), + }; + let raw_payload = serde_json::to_string(&artifact).expect("raw json"); + assert_eq!( + artifact.to_json(), + Err(AnalysisEngineError::InvalidEpisodeMembershipArtifact) + ); + assert_eq!( + EpisodeMembershipArtifact::from_json(&raw_payload), + Err(AnalysisEngineError::InvalidEpisodeMembershipArtifact) + ); +} + #[test] fn future_available_assignments_are_excluded() { let request = request(); From 04a254662fe8bce3f68e6e41a8290ca13d773fdb Mon Sep 17 00:00:00 2001 From: Seongho Bae Date: Wed, 2 Sep 2026 07:08:29 +0900 Subject: [PATCH 3/3] fix(analysis): bound serialized membership census --- crates/analysis_engine/src/episode_membership_artifact.rs | 1 + 1 file changed, 1 insertion(+) diff --git a/crates/analysis_engine/src/episode_membership_artifact.rs b/crates/analysis_engine/src/episode_membership_artifact.rs index b727c9e22..f72e6812c 100644 --- a/crates/analysis_engine/src/episode_membership_artifact.rs +++ b/crates/analysis_engine/src/episode_membership_artifact.rs @@ -156,6 +156,7 @@ impl EpisodeMembershipArtifact { || !valid_identifier(&self.snapshot_id) || KnowledgeCutoff::parse_rfc3339(&self.knowledge_cutoff).is_err() || self.assignment_count < 2 + || self.assignment_count > MAX_EVIDENCE_UNITS as u64 || self.contained_count == 0 || self.escaped_count == 0 || status_sum != Some(self.assignment_count)