Money Demo v2

Money Demo v2 — End-to-End Domain Mode Exercise

One command. Ten steps. Four domains. Full governance loop with before/after metrics.

Overview

The Money Demo v2 is a 10-step pipeline that exercises all four domain modes (IntelOps, FranOps, ReflectionOps, AuthorityOps) plus the cascade engine. It ingests baseline claims, introduces a contradicting delta, detects drift, executes a retcon, propagates cascades, runs coherence checks, and seals the episode.

Run it: make demo-money or python -m enterprise.src.demos.money_demo

Module: enterprise/src/demos/money_demo/pipeline.py

Pipeline Steps

Step	Name	Domain	What Happens
1	LOAD	—	Load baseline (3 claims) + delta (1 contradicting claim) from fixtures
2	INTELOPS INGEST	IntelOps	Ingest all baseline claims via INTEL-F01
3	INTELOPS VALIDATE	IntelOps	Validate all claims via INTEL-F02
4	INTELOPS DELTA	IntelOps	Ingest contradicting claim, detect drift (INTEL-F01 -> F02 -> F03)
5	FRANOPS PROPOSE	FranOps	Propose and bless canon entries (FRAN-F01 -> F02)
6	FRANOPS RETCON	FranOps	Assess and execute retcon on contradicted claim (FRAN-F04 -> F05)
7	REOPS EPISODE	ReflectionOps	Begin episode, evaluate gates, attest non-coercion, seal (RE-F01 -> F04 -> F07 -> F02)
8	CASCADE	Cascade	Propagate retcon event through cross-domain rules
9	COHERENCE	ReflectionOps	Severity scoring (RE-F08) + coherence check (RE-F09)
10	SEAL	—	Compute summary with audit chain verification

Fixture Data

Baseline (fixtures/baseline.json)

Three claims about a fictional ML churn prediction model:

• CLAIM-MONEY-001: Churn model accuracy (92%)
• CLAIM-MONEY-002: Churn model deployment readiness
• CLAIM-MONEY-003: Data pipeline freshness

Delta (fixtures/delta.json)

One contradicting claim:

• CLAIM-MONEY-004: Churn model v3 underperforms on 2026 data (78% vs claimed 92%)

This triggers contradiction detection, drift signals, and a retcon of CLAIM-MONEY-002.

Output Metrics (MoneyDemoResult)

Metric	Description
baselineClaims	Number of baseline claims loaded
deltaClaims	Number of delta claims loaded
driftSignalsTotal	Total drift signals detected across all steps
retconExecuted	Whether the retcon completed successfully
cascadeRulesTriggered	Number of cascade rules that fired
coherenceScore	Final coherence score
episodeSealed	Whether the episode was sealed
auditEntries	Total entries in the hash-chained audit log
elapsedMs	Pipeline execution time in milliseconds

Infrastructure Initialized

The pipeline creates:

• IntelOps, FranOps, ReflectionOps, AuthorityOps domain mode instances
• CascadeEngine with all four domains registered
• MemoryGraph for claim/drift/patch nodes
• DriftSignalCollector for drift ingestion
• EpisodeTracker for episode state management
• CanonWorkflow for canon state machine
• AuditLog for hash-chained non-coercion logging

Deterministic Verification

Every handler call returns a FunctionResult with a replay_hash (SHA-256 of deterministic output fields). The demo can be replayed to verify identical outputs.

CI Integration

tests/test_money_demo_v2.py runs the full pipeline as a regression test:

• Verifies all 10 steps complete
• Checks retcon executed
• Checks episode sealed
• Validates audit chain integrity
• Asserts coherence score is within expected range

Related Pages

• IntelOps — claim lifecycle domain
• FranOps — canon enforcement domain
• ReflectionOps — gate enforcement domain
• AuthorityOps — authority enforcement domain
• Cascade Engine — cross-domain propagation
• Event Contracts — routing table