Secondary technical evidence
Technical verification register.
This interim register records the packaged engineering evidence assembled through 6 August 2026: repository, command, input, artifact, boundary and status. The live interface captures on the public Evidence page are intentionally presented separately while the version- and event-specific capability map is rebuilt. Downloadable register copies: Markdown · JSON.
Public download copies normalize local filesystem path prefixes before publication; analytical values, statuses and test outcomes are otherwise preserved.
EVA (12) KAILA (6) CSUITE (2) STAR (3) EOS/EPL (7) Specialists (3)
EVA (12 items)
EVA-01Executed
- Application
- EVA
- Capability
- Event impact & stakeholder consequences
- Repository / archive
- EVA/backend
- Commit / hash
4f030361- Command / endpoint
pytest engine10/tests/test_2_6_direct_impact_correctness.py engine10/tests/test_2_7_conservation_pair_ecosystem.py- Fixture / input
- GOLDEN_2_6 / HOU_REG_CFPB_ENFORCEMENT_ACTION; buildings v3.1 flood pair
- World version
- us_housing (stakeholderkit) / buildings v3.1
- Model version
- stakeholderkit RegulatoryEnforcementFine/Revenue/ConsumerRemediation; mortgagekit v9
- Seed
- not_applicable
- Execution date
- 2026-08-06
- Claim
- One event produces typed, signed per-stakeholder consequences that conserve to the transfer.
- Boundary
- Synthetic golden ecosystems. Servicer -180,000,000.00 / regulator +100,000,000.00 / taxpayer welfare; flood owner +6,550,000.00 / insurer -6,550,000.00 sum-to-zero.
- Public page
- evidence.html#eva
- Notes
- Captured in the consolidated six-families run.
- Artifact
- eva-six-families-run.txt ↓
EVA-02Executed
- Application
- EVA
- Capability
- Investment Workbench
- Repository / archive
- EVA/backend
- Commit / hash
4f030361- Command / endpoint
pytest portfolio_intel/tests/{test_portfolio_intel_engine,test_impact_tracer,test_no_fabrication_unvalued,test_cross_module_evaluator}.py- Fixture / input
- portfolio_intel/seed_portfolios.py
- World version
- not_applicable
- Model version
- investment_intel valuation models
- Seed
- not_applicable
- Execution date
- 2026-08-06
- Claim
- A portfolio is evaluated against events with per-position attribution and honest UNVALUED.
- Boundary
- Strategy/execution leg source-verified, not separately executed.
- Public page
- evidence.html#eva
- Notes
- Captured in the consolidated six-families run.
- Artifact
- eva-six-families-run.txt ↓
EVA-03Executed
- Application
- EVA
- Capability
- Forensics
- Repository / archive
- EVA/backend
- Commit / hash
4f030361- Command / endpoint
pytest forensics/tests/{test_forensic_engine,test_counterfactual,test_forensic_query,test_loan_intake,test_regulatory_report}.py- Fixture / input
- forensics fixtures
- World version
- not_applicable
- Model version
- not_applicable
- Seed
- not_applicable
- Execution date
- 2026-08-06
- Claim
- Forensic attribution runs backward from an outcome with counterfactuals and a case artifact.
- Boundary
- 3 temporal/reverse-traversal tests error on the archived-engine9 tripwire (migration targets, not defects).
- Public page
- evidence.html#eva
- Notes
- Tripwire-erroring files excluded from the consolidated run; live replacements pass.
- Artifact
- eva-six-families-run.txt ↓
EVA-04Executed
- Application
- EVA
- Capability
- Decision / Persona
- Repository / archive
- EVA/backend
- Commit / hash
4f030361- Command / endpoint
pytest game_theory/tests/ decision/tests/{test_reaction_models,test_decision_objects}.py- Fixture / input
- game_theory + decision fixtures
- World version
- not_applicable
- Model version
- not_applicable
- Seed
- not_applicable
- Execution date
- 2026-08-06
- Claim
- A proposed action is evaluated against reactive actors with a payoff/minimax verdict and a decision record.
- Boundary
- No literal 'backfire verdict' class — verdict = reaction selection. test_decision_engine.py stale on engine9 tripwire.
- Public page
- evidence.html#eva
- Notes
- —
- Artifact
- eva-six-families-run.txt ↓
EVA-05Executed
- Application
- EVA
- Capability
- Cross-ecosystem / systemic
- Repository / archive
- EVA/backend
- Commit / hash
4f030361- Command / endpoint
pytest identity/tests/ grkg/tests/{test_grkg_propagation,test_grkg_engine}.py- Fixture / input
- identity + grkg fixtures
- World version
- not_applicable
- Model version
- not_applicable
- Seed
- not_applicable
- Execution date
- 2026-08-06
- Claim
- Identity links + shared stakeholders/events drive a named cross-ecosystem propagation path with output.
- Boundary
- systemic/reverberations source-verified only; one (1)-snapshot identity test asserts a stale endpoint count.
- Public page
- evidence.html#eva
- Notes
- —
- Artifact
- eva-six-families-run.txt ↓
EVA-06Executed
- Application
- EVA
- Capability
- Model-input governance & refusal
- Repository / archive
- EVA/backend
- Commit / hash
4f030361- Command / endpoint
pytest engine10/tests/{test_f10_typed_input_defaults,test_phase1_resolver}.py portfolio_intel/tests/test_no_fabrication_unvalued.py- Fixture / input
- F10 typed defaults; resolver fixtures
- World version
- not_applicable
- Model version
- not_applicable
- Seed
- not_applicable
- Execution date
- 2026-08-06
- Claim
- Inputs resolve runtime->asset->cited default->MISSING; a missing required input refuses, never silently defaults.
- Boundary
- Status is REAL / PARTIAL / UNVALUED; governed no-substitute behavior.
- Public page
- evidence.html#eva
- Notes
- —
- Artifact
- eva-six-families-run.txt ↓
EVA-07Source verified
- Application
- EVA
- Capability
- 13-stage firing contract
- Repository / archive
- EVA/backend
- Commit / hash
4f030361- Command / endpoint
read engine10/firing.py:156-181 (fire_event; on_stage_complete hook)- Fixture / input
- firing.py
- World version
- not_applicable
- Model version
- engine10
- Seed
- not_applicable
- Execution date
- 2026-08-06
- Claim
- EVA fires an event through a fixed 13-stage pipeline via one fire_event entrypoint.
- Boundary
- Documented stage order from live source; a live 13-stage trace is captured separately (EVA-12).
- Public page
- evidence.html#eva
- Notes
- —
- Artifact
- A1_firing_stage_sequence.txt ↓
EVA-08Executed
- Application
- EVA
- Capability
- Live firing/values regression
- Repository / archive
- EVA/backend
- Commit / hash
4f030361- Command / endpoint
pytest engine10/tests/test_us_housing_v9_permanent_regression.py -v- Fixture / input
- ecosystems/us_housing_v9/runtime_payload_example.json
- World version
- us_housing_v9
- Model version
- engine10
- Seed
- not_applicable
- Execution date
- 2026-08-06
- Claim
- A live firing reproduces the borrower equity decomposition (equity-realization P&L + lien-release balance-sheet swap).
- Boundary
- Regression-test record, NOT a full 13-stage trace. 5 passed. Asserted EXPECTED_EQUITY_REALIZATION=175,111.0244 (post-migration model-derived sale_proceeds 560,000.4411), tolerance +/-0.05; the docstring's 175,110.5833 is a stale derived intermediate from the old hardcoded 560,000 sale price and is historical only.
- Public page
- evidence.html#eva
- Notes
- Full raw pytest capture replaces the earlier curated note.
- Artifact
- eva-firing-regression-full.txt ↓
EVA-09Executed
- Application
- EVA
- Capability
- Hash-locked deterministic replay
- Repository / archive
- EVA/backend
- Commit / hash
4f030361- Command / endpoint
pytest engine10/tests/test_phase8_replay.py -v- Fixture / input
- engine10/tests/fixtures/phase8_replay_baseline.json
- World version
- us_housing_v9
- Model version
- engine10
- Seed
- not_applicable
- Execution date
- 2026-08-06
- Claim
- A firing path replays byte-identically to a recorded trace-hash baseline (12 events locked).
- Boundary
- 4 passed, 1 deprecation warning. Full raw capture incl. summary. Hashes are asserted in-test against the baseline fixture and do not print to stdout; baseline HOME_SALE_PAYOFF trace_event_v2_hash b15d1fad... lives in the fixture.
- Public page
- evidence.html#eva
- Notes
- Replaces the earlier curated extract.
- Artifact
- deterministic-replay-full.txt ↓
EVA-10Source verified
- Application
- EVA
- Capability
- Access surfaces (direct / REST / MCP)
- Repository / archive
- EVA/backend + eva_mcp
- Commit / hash
4f030361- Command / endpoint
engine10/api.py:727 POST /api/v1/firing/fire-event ; eva_mcp/eva_mcp/server.py (9 tools)- Fixture / input
- not_applicable
- World version
- not_applicable
- Model version
- engine10
- Seed
- not_applicable
- Execution date
- 2026-08-06
- Claim
- One canonical fire_event is invocable directly, over REST, and over MCP (connected in Claude Desktop).
- Boundary
- Direct = Executed (EVA-01/08). REST = source-verified, server not booted. MCP = implemented + connected (claude_desktop_config.json), acceptance test not run (permission-denied); no transcript fabricated.
- Public page
- evidence.html#eva
- Notes
- Live REST/MCP runs remain in the gap plan.
- Artifact
- —
EVA-11Observed interface
- Application
- EVA
- Capability
- Operator interface (breadth)
- Repository / archive
- EVA/frontend
- Commit / hash
not_applicable- Command / endpoint
captured screenshots- Fixture / input
- not_applicable
- World version
- not_applicable
- Model version
- not_applicable
- Seed
- not_applicable
- Execution date
- 2026-08-06
- Claim
- EVA is a substantial operator environment with many interface surfaces.
- Boundary
- Interface existence/breadth only; captured 2026-08-04; does not by itself prove every backend path.
- Public page
- evidence.html#eva
- Notes
- Also eva-event-impact.webp, eva-decision.webp.
- Artifact
- assets/screenshots/eva-operator.webp ↓
EVA-12Executed
- Application
- EVA
- Capability
- 13-stage firing trace (live, offline)
- Repository / archive
- EVA/backend
- Commit / hash
4f030361- Command / endpoint
read-only driver attaching on_stage_complete to fire_event (offline PackageLoader+PythonRunner)- Fixture / input
- ecosystems/us_housing_v9/runtime_payload_example.json
- World version
- us_housing_v9
- Model version
- engine10
- Seed
- not_applicable
- Execution date
- 2026-08-06
- Claim
- All 13 firing stages are observed live for one event, with 27 stakeholder impacts and channel_1=4.
- Boundary
- Different payload from the phase8 baseline, so this trace's event_v2_hash (a0400b24...) is NOT the phase8 replay hash. Offline, no server.
- Public page
- evidence.html#eva
- Notes
- Driver lives in scratchpad, not in the EVA repo; capture is raw stdout.
- Artifact
- eva-13stage-trace.txt ↓
KAILA (6 items)
KAILA-01Executed
- Application
- KAILA
- Capability
- Canonical signal + version negotiation (Connect)
- Repository / archive
- KINDYNOS APPS/KAILA/backend
- Commit / hash
not_applicable- Command / endpoint
pytest sdk/tests/test_data_contracts.py sdk/tests/test_version_negotiation.py -v- Fixture / input
- sdk RiskSignal fixtures
- World version
- not_applicable
- Model version
- not_applicable
- Seed
- not_applicable
- Execution date
- 2026-08-06
- Claim
- A canonical RiskSignal validates with a deterministic content hash, and incompatible engine versions are rejected.
- Boundary
- 47 passed (test_canonical_hash, test_incompatible_raises_version_conflict). In-process only.
- Public page
- evidence.html#kaila
- Notes
- —
- Artifact
- kaila-signal-contract-run.txt ↓
KAILA-02Executed
- Application
- KAILA
- Capability
- Artifact registry + hash-chained audit (Connect)
- Repository / archive
- KINDYNOS APPS/KAILA/backend
- Commit / hash
not_applicable- Command / endpoint
pytest registry/tests/test_registry.py -v- Fixture / input
- registry fixtures
- World version
- not_applicable
- Model version
- not_applicable
- Seed
- not_applicable
- Execution date
- 2026-08-06
- Claim
- Artifacts register with content hashes and a verifiable SHA-256 hash-chained audit ledger.
- Boundary
- 81 passed. In-process registry, not a networked service.
- Public page
- evidence.html#kaila
- Notes
- —
- Artifact
- kaila-registry-run.txt ↓
KAILA-03Executed
- Application
- KAILA
- Capability
- Calibration engine (Connect)
- Repository / archive
- KINDYNOS APPS/KAILA/backend
- Commit / hash
not_applicable- Command / endpoint
pytest calibration/tests/test_calibration.py -v- Fixture / input
- seed 0 fixtures
- World version
- not_applicable
- Model version
- Hawkes (spectral stability) / Gaussian copula (PSD) / MLE
- Seed
- 0
- Execution date
- 2026-08-06
- Claim
- Named quantitative calibration runs and produces real values (Hawkes stability, copula PSD projection, MLE).
- Boundary
- 127 passed (test_stability_constraint, test_psd_enforced, test_psd_projection_2x2). Pure-Python.
- Public page
- evidence.html#kaila
- Notes
- —
- Artifact
- kaila-calibration-run.txt ↓
KAILA-04Source verified
- Application
- KAILA
- Capability
- Global freeze governance (Connect)
- Repository / archive
- KINDYNOS APPS/KAILA/backend
- Commit / hash
not_applicable- Command / endpoint
freeze/global_freeze.py (in-process over temp SQLite, prior reader demo)- Fixture / input
- not_applicable
- World version
- not_applicable
- Model version
- not_applicable
- Seed
- not_applicable
- Execution date
- 2026-08-06
- Claim
- A cross-app freeze snapshot enforces and raises on missing declarations.
- Boundary
- Demonstrated in a prior in-process reader demo (FreezeViolationError); not separately packaged this pass.
- Public page
- evidence.html#kaila
- Notes
- qualifier: prior run, in-process.
- Artifact
- —
KAILA-05Source verified
- Application
- KAILA
- Capability
- Mediation / route resolution (Connect)
- Repository / archive
- KINDYNOS APPS/KAILA/backend
- Commit / hash
not_applicable- Command / endpoint
tests/test_phase2_mediation.py (source-verified; prior reader run 83 passed)- Fixture / input
- not_applicable
- World version
- not_applicable
- Model version
- not_applicable
- Seed
- not_applicable
- Execution date
- 2026-08-06
- Claim
- A signal's route resolves and hops are hash-chained to a finalized chain.
- Boundary
- Recorded routing INTENT + in-process hash-chaining, NOT network transport delivery. KAILA does no transport in this codebase.
- Public page
- evidence.html#kaila
- Notes
- qualifier: in-process; delivery status is caller-supplied, not executed transport.
- Artifact
- —
KAILA-06Test exercised
- Application
- KAILA
- Capability
- Mediation spine (Connect)
- Repository / archive
- KINDYNOS APPS/KAILA/backend
- Commit / hash
not_applicable- Command / endpoint
pytest tests/test_phase2_mediation.py (prior reader: 83 passed)- Fixture / input
- not_applicable
- World version
- not_applicable
- Model version
- not_applicable
- Seed
- not_applicable
- Execution date
- 2026-08-06
- Claim
- Route resolution + hop hashing + chain finalization pass as a suite.
- Boundary
- Not re-captured this pass; see KAILA-05 boundary on transport.
- Public page
- evidence.html#kaila
- Notes
- qualifier: prior run.
- Artifact
- —
CSUITE (2 items)
CSUITE-01Executed
- Application
- CSUITE
- Capability
- Deterministic rehearsal spine (Rehearse)
- Repository / archive
- KINDYNOS/OTHER APPS/CSUITE/backend
- Commit / hash
not_applicable- Command / endpoint
pytest execution/tests -v- Fixture / input
- execution fixtures
- World version
- not_applicable
- Model version
- not_applicable
- Seed
- not_applicable
- Execution date
- 2026-08-06
- Claim
- A rehearsal freezes state, forks branches, applies interventions, and verifies an append-only hash-chain replay.
- Boundary
- 49 passed (test_fork, test_verify_replay, test_get_final_hash, test_immutability_fork_replay, test_persist_intervention). Agent reactions are deterministic KAILA sandbox stubs, not a live LLM.
- Public page
- evidence.html#csuite
- Notes
- —
- Artifact
- csuite-execution-suite-run.txt ↓
CSUITE-02Source verified
- Application
- CSUITE
- Capability
- Scored rehearsal session (Rehearse)
- Repository / archive
- KINDYNOS/OTHER APPS/CSUITE/backend
- Commit / hash
not_applicable- Command / endpoint
read csuite_session.db (ro)- Fixture / input
- not_applicable
- World version
- not_applicable
- Model version
- not_applicable
- Seed
- 99
- Execution date
- 2026-08-06
- Claim
- A completed rehearsal session carries a component-weighted score that reconciles.
- Boundary
- SSN_49443fc2 MARKET_SPREAD_SHOCK overall_score 0.725 = 0.4x0.8 + 0.3x0.7 + 0.3x0.65. Prior run, not re-executed; top-level session route persisted 0 branches.
- Public page
- evidence.html#csuite
- Notes
- qualifier: prior run (database read).
- Artifact
- —
STAR (3 items)
STAR-01Executed
- Application
- STAR
- Capability
- v4 governed media pipeline (Communicate)
- Repository / archive
- KINDYNOS/OTHER APPS/STAR/backend/v4
- Commit / hash
not_applicable- Command / endpoint
pytest v4/tests/ -v- Fixture / input
- EVA_REPORT canonical bundle
- World version
- not_applicable
- Model version
- engine_registry_v2 4.0.0
- Seed
- not_applicable
- Execution date
- 2026-08-06
- Claim
- A governed media pipeline deterministically plans scenes, assigns job identity, assembles metadata, records a 7-stage replay ledger.
- Boundary
- 386 passed. Render is PLANNING-ONLY — no external provider invoked; render outputs are placeholder hashes.
- Public page
- evidence.html#star
- Notes
- —
- Artifact
- star-v4-suite-run.txt ↓
STAR-02Executed
- Application
- STAR
- Capability
- Determinism certification gate (Communicate)
- Repository / archive
- KINDYNOS/OTHER APPS/STAR/backend/v4
- Commit / hash
not_applicable- Command / endpoint
pytest v4/tests/test_certification.py v4/tests/test_determinism.py -v- Fixture / input
- EVA_REPORT bundle
- World version
- not_applicable
- Model version
- engine_registry_v2 4.0.0
- Seed
- not_applicable
- Execution date
- 2026-08-06
- Claim
- Running the institutional pipeline twice yields a byte-identical cumulative_hash (asserted in-test).
- Boundary
- 75 passed (test_cert_identical_input_identical_hashes, test_cert_cumulative_hash_stable, test_cert_chain_integrity_N_runs, test_cross_bundle_determinism).
- Public page
- evidence.html#star
- Notes
- —
- Artifact
- star-determinism-gate.txt ↓
STAR-03Observed interface
- Application
- STAR
- Capability
- Pre-rendered exports (Communicate)
- Repository / archive
- KINDYNOS/OTHER APPS/STAR
- Commit / hash
not_applicable- Command / endpoint
ls STAR MEDIA/exports/vega_export_*.mp4- Fixture / input
- not_applicable
- World version
- not_applicable
- Model version
- not_applicable
- Seed
- not_applicable
- Execution date
- 2026-08-06
- Claim
- Prior rendered outputs exist on disk.
- Boundary
- 11 vega_export_*.mp4 from a separate non-v4 render path; existence only — NOT outputs of the v4 deterministic pipeline; no live render performed.
- Public page
- evidence.html#star
- Notes
- qualifier: prior run, existence only.
- Artifact
- —
EOS/EPL (7 items)
EOS-01Executed
- Application
- EOS/EPL
- Capability
- Named refusal (backend unavailable)
- Repository / archive
- EOS v0.24
- Commit / hash
6b9ae85- Command / endpoint
eos run sequential_crash- Fixture / input
- sequential_crash.epl
- World version
- not_applicable
- Model version
- not_applicable
- Seed
- not_applicable
- Execution date
- 2026-08-06
- Claim
- A backend-unavailable path returns UNVALUED, publishes no effects, commits no state.
- Boundary
- Two events UNVALUED (BACKEND_UNAVAILABLE), state v0->v0, exit 3. Bounded EOS path.
- Public page
- evidence.html#eos
- Notes
- —
- Artifact
- named-refusal-run.txt ↓
EOS-02Executed
- Application
- EOS/EPL
- Capability
- State carry-forward
- Repository / archive
- EOS v0.24
- Commit / hash
6b9ae85- Command / endpoint
eos run mortgage_rate_path- Fixture / input
- mortgage_rate_path.epl
- World version
- not_applicable
- Model version
- not_applicable
- Seed
- not_applicable
- Execution date
- 2026-08-06
- Claim
- State written by one event is read by the next.
- Boundary
- 3 sequential events; balance/cum_prepay carry forward; example model.
- Public page
- evidence.html#eos
- Notes
- —
- Artifact
- mortgage-path-run.txt ↓
EOS-03Executed
- Application
- EOS/EPL
- Capability
- Domain-blind kernel
- Repository / archive
- EOS v0.24
- Commit / hash
6b9ae85- Command / endpoint
python3 -m conformance.tests.test_blindness- Fixture / input
- authored worlds
- World version
- not_applicable
- Model version
- not_applicable
- Seed
- not_applicable
- Execution date
- 2026-08-06
- Claim
- The execution core contains none of the domains it runs.
- Boundary
- 6 sub-proofs, exit 0; small authored worlds.
- Public page
- evidence.html#eos
- Notes
- —
- Artifact
- eos_domain_blindness.txt ↓
EOS-04Executed
- Application
- EOS/EPL
- Capability
- Portfolio reconciliation
- Repository / archive
- EOS v0.24
- Commit / hash
6b9ae85- Command / endpoint
eos run book_50- Fixture / input
- book_50.epl
- World version
- not_applicable
- Model version
- not_applicable
- Seed
- not_applicable
- Execution date
- 2026-08-06
- Claim
- Per-position results sum exactly to the book total.
- Boundary
- BOOK delta +433,004.662908, 50 VALUED; synthetic book.
- Public page
- evidence.html#eos
- Notes
- —
- Artifact
- portfolio-50-run.txt ↓
EOS-05Source verified
- Application
- EOS/EPL
- Capability
- EPL source example
- Repository / archive
- EOS v0.24
- Commit / hash
6b9ae85- Command / endpoint
not_applicable- Fixture / input
- not_applicable
- World version
- not_applicable
- Model version
- not_applicable
- Seed
- not_applicable
- Execution date
- 2026-08-06
- Claim
- Declarative EPL source for a sequential-crash example world.
- Boundary
- Source example; not an execution record.
- Public page
- evidence.html#eos
- Notes
- —
- Artifact
- sequential_crash.epl ↓
EOS-06Test exercised
- Application
- EOS/EPL
- Capability
- EOS full suite
- Repository / archive
- EOS v0.24
- Commit / hash
6b9ae85- Command / endpoint
python3 -m pytest- Fixture / input
- not_applicable
- World version
- not_applicable
- Model version
- not_applicable
- Seed
- not_applicable
- Execution date
- 2026-08-06
- Claim
- The EOS test suite passes.
- Boundary
- 29 passed in 10.84s (blindness proof separate).
- Public page
- evidence.html#eos
- Notes
- —
- Artifact
- B2_eos_full_suite.txt ↓
EOS-07Source verified
- Application
- EOS/EPL
- Capability
- Run provenance
- Repository / archive
- EOS v0.24
- Commit / hash
6b9ae85- Command / endpoint
not_applicable- Fixture / input
- not_applicable
- World version
- not_applicable
- Model version
- not_applicable
- Seed
- not_applicable
- Execution date
- 2026-08-06
- Claim
- The EOS output files are unmodified terminal output, re-verified.
- Boundary
- Provenance doc; re-verified 3 Aug 2026 EOS v0.24.
- Public page
- evidence.html#eos
- Notes
- —
- Artifact
- PROVENANCE.txt ↓
Specialists (3 items)
SPEC-01Test exercised
- Application
- Specialist · CRIS
- Capability
- Cyber risk
- Repository / archive
- KINDYNOS_Risk_Apps 2 / CRIS cris_unified 2
- Commit / hash
not_applicable- Command / endpoint
pytest tests/test_cris_signal_contract.py -v ; verbatim PoissonCyberModel fit- Fixture / input
- synthetic 12-month incident series [2,4,3,5,1,6,3,4,7,2,5,4]
- World version
- not_applicable
- Model version
- risk_analytics_core.py PoissonCyberModel (MLE + scipy.stats.poisson)
- Seed
- not_applicable
- Execution date
- 2026-08-06
- Claim
- The canonical cyber RiskSignal contract passes (deterministic hashing + severity mapping); the named Poisson model computes real frequency estimates.
- Boundary
- 17 passed (signal envelope, NOT a calibrated cyber model). Poisson fit lambda 3.8333/mo, E[3mo] 11.5, P(=5|3mo) 0.0169793638.
- Public page
- evidence.html#specialists
- Notes
- model function exercised (Poisson run under a scipy interpreter — CRIS's own venv lacks scipy and the module also needs pandas), NOT integrated application path.
- Artifact
- cris-contract-run.txt ↓
SPEC-02Executed
- Application
- Specialist · DEFIANT
- Capability
- DeFi risk
- Repository / archive
- KINDYNOS_Risk_Apps 2 / DEFIANT backend
- Commit / hash
not_applicable- Command / endpoint
run DefiMonteCarloEngine.run_portfolio_simulation (real backend modules, unmodified)- Fixture / input
- 3-asset ETH/UNI/USDC @ 1,000,000 each, 1000 sims, 30-day horizon
- World version
- not_applicable
- Model version
- defi_monte_carlo.py + defi_risk_engine.py (numpy 2.4.3 / scipy 1.17.1)
- Seed
- 12345
- Execution date
- 2026-08-06
- Claim
- A DeFi portfolio VaR engine runs with Cholesky-correlated shocks and beta-distribution losses, producing VaR percentiles.
- Boundary
- expected_loss 359,695.52; VaR95 508,080.82; VaR99 555,375.34; cholesky_decomposition + beta_distribution; reproducible on same seed.
- Public page
- evidence.html#specialists
- Notes
- research prototype executed: 0 tests, config defaults, synthetic uncalibrated inputs; NOT integrated application path.
- Artifact
- defiant-montecarlo-run.txt ↓
SPEC-03Executed
- Application
- Specialist · USRISK
- Capability
- US economic risk
- Repository / archive
- KINDYNOS_Risk_Apps 2 / USRISK hawkes_full_app_mc_with_docs_v2
- Commit / hash
not_applicable- Command / endpoint
run simulate_copula_hawkes (real hawkes.py + factors.py, unmodified)- Fixture / input
- 3 nodes, 2 Student-t copula factors (df 5), 2000 trials
- World version
- not_applicable
- Model version
- hawkes.py + factors.py (numpy 2.4.2)
- Seed
- 12345
- Execution date
- 2026-08-06
- Claim
- A multivariate self-exciting Hawkes process driven by Student-t copula shocks computes aggregate tail risk, reproducibly.
- Boundary
- branching_ratio 0.16 (subcritical); agg_VaR99 35.0; agg_TVaR99 38.8; bitwise-identical on same-seed re-run.
- Public page
- evidence.html#specialists
- Notes
- research prototype executed: 0 tests, model function only (the Flask app path is non-functional — app.py imports symbols this model version does not provide); synthetic params.
- Artifact
- usrisk-hawkes-run.txt ↓