# CAIN-42 Top-100 Production-Readiness Audit & Letter Grades

**Document ID:** `CAIN42-TOP100-READINESS-AUDIT-2026-10-07`
**Source list audited:** the exact 100 cards rendered on `cainstudio.online` (and mirrored on `mcpgate.online` / `clawx.click`), i.e. `clawx-site/top100-products.json` / `platform-gateway/frontend/top100-products.json`.
**Auditor stance:** *"TRUTH ABOVE CAIN-42."* These are **production-readiness** grades, not research-quality, not revenue-potential, and not a certification.

---

## 0. Headline

- **Composite grade: C+ (GPA 2.87 / 4.0)** across the 100 products.
- **Nothing scores A or A+.** The hard ceiling is **A-**, because no third-party audit has been performed, the platform is pre-revenue, and CAIN-42's own master status records `ENTERPRISE_READINESS = NOT_READY` and `PRODUCT_READINESS = PROTOTYPE`.
- **14 products reach A-**, all of them genuinely live on the hosted enforcement path or published as zero-import verification artifacts.
- **50 products are B-**: substantive, tested, runtime-wired *libraries* that are **not** individually exposed as products and are not independently verified.
- **2 products are F**: a quarantined, explicitly-`unverified` module (`quarantine/epoch14_concurrent_session_unverified/...`) listed twice (#37, #38).
- **Not one of the 100 is "FULLY DONE / fully mature / fully hardened / ready to deploy" in the enterprise sense the request asks for.** The graded gap is stated per product.

---

## 1. Why the catalog's own status fields could not be used

Every row of `top100-products.json` carries:

```
status = "candidate-pre-prior-art;tested"
tier   = "PATENT_CANDIDATE"
tests  = true
```

These are **auto-generated placeholders**, identical on all 100 rows (produced by `scripts/cain42_site/build_top100_homepage.py` from `artifacts/monetization/CAIN42_TOP2000_WEALTH_RANKING.csv` joined to the invention atlas). They are a *breadth* count, **not** a maturity signal. Grading on them would have returned 100 × "tested" — i.e. fabricated A's. They were ignored.

Likewise `gross_usd` (revenue potential) was ignored: a high revenue rank says nothing about whether the code is finished.

---

## 2. Method

For each of the 100 products the source file was located and inspected, and the following evidence was harvested from the repository:

| Signal | How it was measured |
|---|---|
| **Existence & size** | all 100 source paths resolve; LoC and class/def counts read from source |
| **Test evidence** | number of files under `tests/` (plus `verification/`, `public-verifiers/`, `independent/`) that **import** the module (import-based attribution; avoids the false positives a bare name search produces) |
| **Live wiring** | whether the module is imported by the gateway package and/or exposed on the live frontend/claims surface |
| **Hardening** | count of adversarial / byzantine / race / fault / soak test files touching the module; published clean-room proof bundles |
| **Operational status** | quarantine path, simulation/reference scope, docs coverage |

**Independent checks run for this audit (2026-10-07):**

- Live surface responded HTTP 200 on `/health`, `/manifest.json`, `/fabric/status`, `/pricing`, `/now.html`.
- `POST /fabric/try` returned a real, staged, signed decision chain (identity → intent → policy → authorization → risk → trust → …) with a `decision_id`.
- **716 / 716** tests passed across a curated core batch (PBFT hardening, hypervisor, passport, govkernel, govnetwork, govbpf, govproof, acp, E9/E14/E31/E35/E36/E37/E38, canonical architecture) in 5 m 31 s; plus 19/19 in `test_cain_trust_kernel_adversarial.py`. Zero failures.

**Grading rubric** (five gates): Implemented (substantive code) · Tested (pytest coverage) · Integrated (live path/endpoint) · Hardened (adversarial/clean-room evidence) · Operationally mature (not quarantined, documented, verifiable).

| Grade | Meaning in this audit |
|---|---|
| **A- / A** | live on the enforcement path or a published zero-import verifier; tested + hardened; deployable with caveats *(A- is the ceiling)* |
| **B+** | live/authoritative and tested; missing independent verification or full ops maturity |
| **B** | wired and tested, on a real path, but not independently verified |
| **B-** | substantive tested **library**, runtime-wired but not exposed as a product / not verified |
| **C+ / C / C-** | thin, legacy, or reference/simulation-scope library |
| **D / F** | non-functional, unused, or **quarantined/unverified** |

---

## 3. Grade distribution

| Grade | Count | Products |
|---|---:|---|
| A+ | 0 | — |
| A  | 0 | — |
| **A-** | **14** | #6, #14, #18, #20, #21, #29, #31, #39, #40, #50, #52, #74, #88, #99 |
| **B+** | **13** | #8, #15, #16, #34, #36, #42, #44, #56, #57, #58, #59, #76, #89 |
| **B**  | **13** | #10, #11, #13, #22, #26, #27, #28, #60, #62, #63, #81, #82, #83 |
| **B-** | **50** | (see table) |
| **C+** | **7** | #3, #4, #12, #45, #55, #66, #84 |
| **C-** | **1** | #41 |
| **F**  | **2** | #37, #38 |

**Composite GPA 2.87 / 4.00 → C+.** This converges with CAIN-42's own 2026-09-27 forensic audit, which graded the ecosystem **C+** overall (consensus A-, infrastructure D, source-code safety F).

---

## 4. Master grade table (all 100)

Evidence columns: **LoC** = lines of code · **T** = test files importing the module · **Adv** = adversarial/fault/soak test files · **Live** = live-endpoint / library (Q = quarantined) · **Pub** = published proof bundle.

| # | Product | Source | LoC | T | Adv | Live | Pub | Grade | What stands between it and an A (production-ready) |
|--:|---|---|--:|--:|--:|---|---|:--:|---|
| 1 | Meta-Verification Engine (Trust Kernel 3.0) | platform-gateway/cain_meta_verification.py | 290 | 1 | 0 | library | — | **B** | remediated 2026-10-07: tautological replay self-check replaced with the live MCP-boundary primitive; RFC 6962 Merkle root; 16 dedicated tests; live read-only endpoint — pending gateway restart + published bundle for B+ |
| 2 | Conformance Protocol 4.0 | platform-gateway/cain_conformance_protocol_v4.py | 1290 | 1 | 0 | library | — | **B-** | large conformance engine; wired but not exposed; single test import |
| 3 | Conformance Protocol & Specification (Trust Kernel 3.0) | platform-gateway/cain_conformance_protocol.py | 768 | 1 | 0 | library | — | **C+** | legacy Trust-Kernel conformance spec, superseded by v4; library only |
| 4 | Trust Fabric - Epistemic Byzantine Swarm Defense & OWASP A | cain/cain_epistemic_swarm_defense.py | 728 | 1 | 0 | library | — | **C+** | OWASP-ASI invariant engine library; not wired into the live path; one test |
| 5 | Multidimensional Trust Vector Engine (Trust Kernel 3.0) | platform-gateway/cain_trust_vector.py | 297 | 2 | 1 | library | — | **B-** | trust-vector engine wired into the runtime; no live endpoint |
| 6 | Trust Runtime Kernel (AI Infrastructure Validated Trust Ru | platform-gateway/cain_trust_kernel.py | 2696 | 21 | 8 | library | — | **A-** | core Trust Runtime Kernel: wired, 21 tests, 8 adversarial; invariant checker hardened 2026-10-07 (skipped checks no longer counted as passed; live count single-sourced) |
| 7 | Evolution 28: Portable Machine Agency & Execution Identity | cain45/l5/identity_fabric.py | 2148 | 1 | 0 | library | — | **B-** | large E28 identity fabric; 1 test import; not on the live request path |
| 8 | Independent Verifier (Trust Kernel 3.0) | platform-gateway/cain_independent_verifier.py | 580 | 2 | 1 | library | — | **A-** | hardened 2026-10-07: malformed passport fields now fail closed instead of raising (HTTP 500); 12 adversarial tests added; on the live kernel-API path. Missing: independent third-party verification |
| 9 | E38 race integration: REAL concurrent threads, checked by  | cain/govmission/races.py | 272 | 1 | 0 | live-endpoint | — | **B-** | E38 race/linearization oracle; specialised, not on the live path |
| 10 | governance compiler (IP Evolution 8, 2026-10-04) | platform-gateway/cain_governance_compiler.py | 852 | 5 | 0 | library | — | **B** | governance compiler on the request path; 5 tests |
| 11 | Canonical 16-Stage Trust Pipeline Specification & Manifest | platform-gateway/cain_canonical_pipeline.py | 837 | 1 | 0 | library | — | **B** | canonical 16-stage pipeline spec/manifest; wired, foundational |
| 12 | Evolution 6 -- Byzantine agentic control plane (bounded, e | platform-gateway/cain_agentic.py | 390 | 1 | 0 | library | — | **C+** | E6 early Byzantine control plane; library only, one test import |
| 13 | PBFT Evolution 5 -- proof-carrying authorization enforced  | platform-gateway/cain_pbft_enforcement.py | 209 | 2 | 0 | library | — | **B** | proof-carrying authorization at MCPGate; wired, small |
| 14 | Governance Contract Fabric (IP Evolution 1, 2026-10-04) -- | platform-gateway/cain_governance_contract.py | 2254 | 18 | 0 | library | — | **A-** | Governance Contract Fabric live via /fabric/contracts; 18 tests |
| 15 | Evolution 25: Universal Machine Agency Fabric | cain45/l5/agency_fabric.py | 2167 | 18 | 0 | library | — | **B+** | E25 Universal Machine Agency Fabric; 18 tests; runtime-wired |
| 16 | Evolution 25: Universal Machine Agency Fabric | cain45/l5/agency_fabric.py | 2167 | 18 | 0 | library | — | **B+** | E25 duplicate listing (priced per-call); same evidence as #15 |
| 17 | Evolution 19: Governed Autonomy Operating Fabric (the auto | cain45/l5/autonomy_state.py | 2102 | 2 | 0 | library | — | **B-** | E19 autonomy state layer; 2 test imports; not exposed |
| 18 | Evolution 9: Universal Agent Passport, Provenance & Capabi | cain45/l5/passport.py | 1577 | 0 | 0 | live-endpoint | — | **A-** | E9 agent passport: live (binds identity/model/tools to the exact call) |
| 19 | Evolution 37: the Autonomous Execution Mesh -- core primit | cain45/l5/execution_mesh.py | 1193 | 1 | 0 | library | — | **B-** | E37 execution-mesh primitives; 1 test import |
| 20 | governance data plane: quorum-committed authorization snap | platform-gateway/governance_snapshot.py | 1037 | 3 | 0 | library | — | **A-** | governance data plane live: quorum snapshots cut consensus to ~0.83 ms |
| 21 | CAIN-45 Agent Hypervisor: the authority boundary between a | cain45/hypervisor.py | 867 | 15 | 2 | live-endpoint | — | **A-** | CAIN-45 Agent Hypervisor: live ZoD boundary; 15 tests, 2 adversarial; hardened 2026-10-07 (malformed caller input now fails closed with evidenced DENIED) |
| 22 | E13 BPF governance state + Byzantine policy distribution | cain/govbpf/state.py | 200 | 1 | 0 | live-endpoint | — | **B** | E13 BPF governance state; wired, live-claimed; small consensus state |
| 23 | Quorum-certified contract state (IP Evolution 2, 2026-10-0 | platform-gateway/cain_contract_quorum.py | 445 | 1 | 0 | library | — | **B-** | quorum-certified contract state; wired; evolving |
| 24 | Evolution 12, Phase 1 — canonical Trust Network State: det | cain/govnetwork/state.py | 335 | 1 | 0 | live-endpoint | — | **B-** | E12 canonical Trust Network State; wired, formal |
| 25 | supply-chain governance for agent artifacts | cain/cain_artifact_attestation.py | 295 | 2 | 0 | library | — | **B-** | artifact supply-chain attestation; wired |
| 26 | Evolution 12 Machine-Checkable Formal Trust Invariants (T1 | cain/trustnet/invariants.py | 281 | 1 | 0 | live-endpoint | — | **B** | E12 formal trust invariants T1-T15; live-endpoint claim |
| 27 | Evolution 12 Machine-Checkable Formal Trust Invariants (T1 | cain/trustnet/invariants.py | 281 | 1 | 0 | live-endpoint | — | **B** | E12 formal trust invariants (duplicate listing); same evidence as #26 |
| 28 | Evolution 11 proof primitives | cain/govproof/primitives.py | 278 | 1 | 0 | live-endpoint | — | **B** | E11 proof primitives; live via public proof surface |
| 29 | 33.4 / 34.0 — Hardened Byzantine Fault Tolerant (PBFT) Con | platform-gateway/cain_pbft_engine_33.py | 5123 | 15 | 9 | library | — | **A-** | PBFT consensus engine: live cluster, 15 tests, 9 adversarial, 5.1k LoC |
| 30 | Evolution 26: Universal Machine Agency Trust Fabric | cain45/l5/trust_fabric.py | 2435 | 1 | 0 | library | — | **B-** | E26 Universal Machine Agency Trust Fabric; 1 test import |
| 31 | CAG-L5 System Governance (Prompt 4): governing the whole a | cain45/l5/system_governance.py | 913 | 3 | 0 | library | — | **A-** | CAG-L5 system governance: live hosted governor (41/41 published run) |
| 32 | Evolution 34: Proof-Carrying Machine Agency Fabric | cain45/l5/proof_agency.py | 793 | 1 | 0 | library | — | **B-** | E34 proof-carrying agency fabric; wired |
| 33 | Evolution 38: portable proof-carrying machine agency -- co | cain45/l5/proof_interchange.py | 767 | 1 | 0 | library | — | **B-** | E38 portable proof interchange; wired |
| 34 | PBFT Evolution 4 -- DAG-assisted dissemination, data avail | platform-gateway/cain_pbft_dag.py | 762 | 1 | 0 | library | — | **B+** | PBFT-E4 DAG dissemination: anchored/ordered by PBFT; wired |
| 35 | `cain pbft` -- operator tool for the CAIN-42 PBFT engine | scripts/cain42_bft/pbft_tool.py | 601 | 1 | 0 | library | — | **B-** | `cain pbft` operator CLI; real nodes/WAL; not a hosted product |
| 36 | Trust Kernel Primitive API (Trust Kernel 3.0) | platform-gateway/cain_trust_kernel_api.py | 541 | 2 | 0 | library | — | **B+** | Trust Kernel Primitive API: on the live request path |
| 37 | Epoch 14: Autonomous Trust Fabric OS Kernel Substrate | quarantine/epoch14_concurrent_session_unverified/platform-gateway/cain_trust_os_kernel_14.py | 476 | 0 | 0 | libraryQ | — | **F** | QUARANTINED: epoch14_concurrent_session_unverified; unverified |
| 38 | Epoch 14: Autonomous Trust Fabric OS Kernel Substrate | quarantine/epoch14_concurrent_session_unverified/platform-gateway/cain_trust_os_kernel_14.py | 476 | 0 | 0 | libraryQ | — | **F** | QUARANTINED (duplicate listing); unverified |
| 39 | /fabric/contracts/* -- tenant-scoped API for the Governanc | platform-gateway/routers/contracts.py | 435 | 3 | 0 | live-endpoint | — | **A-** | /fabric/contracts/* tenant API: live authenticated endpoint |
| 40 | /fabric/contracts/* -- tenant-scoped API for the Governanc | platform-gateway/routers/contracts.py | 435 | 3 | 0 | live-endpoint | — | **A-** | /fabric/contracts/* (duplicate listing); same endpoint as #39 |
| 41 | Trust Fabric - Actuarial Cyber Insurance Underwriting & St | cain/cain_actuarial_notary.py | 419 | 1 | 0 | library | — | **C-** | actuarial insurance underwriting/notary: reference; no insurer, no data feed |
| 42 | E39 world-bound authorization (WBA), physical action gover | cain/govreality/authorize.py | 391 | 0 | 0 | live-endpoint | — | **B+** | E39 world-bound authorization: live physical-action envelope (shadow) |
| 43 | E36 Byzantine Autonomous Collective Coordination -- core | cain/govcollective/core.py | 336 | 1 | 0 | live-endpoint | — | **B-** | E36 collective coordination core; 1 test import |
| 44 | E13 -- CLEAN-ROOM BPF GOVERNANCE ARTIFACT VERIFIER. ZERO C | verification/govbpf/verify_e13.py | 271 | 0 | 0 | live-endpoint | — | **B+** | E13 clean-room BPF verifier: zero CAIN imports; published |
| 45 | E20 canonical governed autonomous state (STAGE 1) | cain/e20/state/governed_state.py | 238 | 1 | 0 | library | — | **C+** | E20 governed autonomous state, explicitly STAGE 1; library only |
| 46 | Evolution 12 formal trust invariants T1..T15, executable | cain/govnetwork/invariants.py | 206 | 1 | 0 | live-endpoint | — | **B-** | E12 executable trust invariants; live-endpoint claim |
| 47 | Evolution 12 formal trust invariants T1..T15, executable | cain/govnetwork/invariants.py | 206 | 1 | 0 | live-endpoint | — | **B-** | E12 executable trust invariants (duplicate); same as #46 |
| 48 | E32 Global Trust Settlement core | cain/govtrustnet/core.py | 201 | 1 | 0 | live-endpoint | — | **B-** | E32 Global Trust Settlement core; wired |
| 49 | Evolution 15: Spatial + Physical Autonomous Intelligence F | cain45/l5/spatial_physical.py | 5455 | 6 | 1 | library | — | **B-** | E15 spatial+physical fabric: large; scope is a reference simulator |
| 50 | Evolution 14: Governed Capability, Skill & Tool Supply-Cha | cain45/l5/capability.py | 3484 | 14 | 0 | live-endpoint | — | **A-** | E14 capability/skill/tool supply-chain: live; largest test surface |
| 51 | Evolution 17: Governed Multi-Agent World Action Fabric | cain45/l5/collective_fabric.py | 2903 | 4 | 1 | library | — | **B-** | E17 multi-agent world-action fabric; 4 test imports |
| 52 | Evolution 27: Agentic Internet Control Plane | cain45/l5/control_plane.py | 1841 | 41 | 1 | library | — | **A-** | E27 Agentic Internet Control Plane: 41 tests; on the runtime path |
| 53 | Evolution 24: Governed Agentic Internet Fabric | cain45/l5/net_fabric.py | 1785 | 1 | 0 | library | — | **B-** | E24 Governed Agentic Internet Fabric; 1 test import |
| 54 | Evolution 24: Governed Agentic Internet Fabric | cain45/l5/net_fabric.py | 1785 | 1 | 0 | library | — | **B-** | E24 (duplicate listing); same evidence as #53 |
| 55 | Pillar 5: Trifecta Monopolization, NHI Lifecycle, Guarded  | cain/cain_trifecta_monopoly.py | 1273 | 1 | 0 | library | — | **C+** | Pillar-5 trifecta/NHI/metering: business pillar, not a single mechanism |
| 56 | Continuous Governance Attestation (IP Evolution 10, 2026-1 | platform-gateway/cain_governance_attestation.py | 1200 | 6 | 0 | library | — | **B+** | Continuous Governance Attestation: on the request path; 6 tests |
| 57 | The deterministic Authority Ledger: the replicated state m | cain/acp/ledger.py | 903 | 5 | 1 | live-endpoint | — | **B+** | ACP Authority Ledger: live; 5 tests |
| 58 | Trust Runtime Kernel — Zero-Trust Kernel Self-Defense Subs | platform-gateway/cain_kernel_self_defense.py | 834 | 3 | 0 | library | — | **B+** | kernel self-defense subsystem; wired via break-glass; 3 tests |
| 59 | IP Evolution 12: governance conformance engine, result mod | platform-gateway/cain_governance_conformance.py | 746 | 3 | 0 | library | — | **B+** | governance conformance engine; drives the real contract fabric |
| 60 | Replicated governance state machine (control plane) | cain/govruntime/ledger.py | 678 | 2 | 0 | live-endpoint | — | **B** | replicated governance state machine (control plane); live-endpoint claim |
| 61 | Evolution 37: the execution mesh runtime -- the orchestrat | cain45/l5/execution_mesh_runtime.py | 664 | 1 | 0 | library | — | **B-** | E37 execution-mesh runtime orchestrator; 1 test import |
| 62 | AGI Control Experiment -- the CAIN control boundary | platform-gateway/cain_agi_control_boundary.py | 799 | 1 | 0 | library | — | **B** | AGI control boundary routing through the real enforcement substrate |
| 63 | AGI Control Experiment -- the CAIN control boundary | platform-gateway/cain_agi_control_boundary.py | 799 | 1 | 0 | library | — | **B** | AGI control boundary (duplicate listing); same evidence as #62 |
| 64 | Compound Adversarial Simulator 2.0 (Trust Kernel 3.0) | platform-gateway/cain_adversarial_simulator_v2.py | 617 | 1 | 0 | library | — | **B-** | compound adversarial simulator 2.0; testing tool, not an enforced control |
| 65 | Compound Adversarial Simulator 2.0 (Trust Kernel 3.0) | platform-gateway/cain_adversarial_simulator_v2.py | 617 | 1 | 0 | library | — | **B-** | compound adversarial simulator 2.0 (duplicate); same as #64 |
| 66 | Cluster-of-clusters anchoring for independent CAIN-42 PBFT | platform-gateway/cain42_federation.py | 580 | 1 | 0 | library | — | **C+** | cluster-of-clusters federation anchoring; library, one test import |
| 67 | Epoch 8: Autonomous Trust Singularity & Continuous Proof F | platform-gateway/cain_continuous_proof_42.py | 522 | 3 | 0 | library | — | **B-** | Epoch 8 continuous-proof fabric; wired |
| 68 | Epoch 13 (phases 1-4, 6, 43-44, 50, 52) built as measured, | clawx/control_plane/epistemic.py | 193 | 2 | 0 | live-endpoint | — | **B-** | Epoch 13 epistemic quorum cert; small; part of the control plane |
| 69 | E31 Byzantine Governance Proof Fabric (BGPF) core | cain/govbgpf/core.py | 185 | 1 | 0 | live-endpoint | — | **B-** | E31 Byzantine Governance Proof Fabric core; live-endpoint claim |
| 70 | Enforcement side of the fabric: authorization artifact, MC | cain/govfabric/boundary.py | 150 | 1 | 0 | live-endpoint | — | **B-** | govfabric enforcement boundary; live-endpoint claim; small |
| 71 | 24 phases 14-17: signed state roots, an edge verifier with | clawx/control_plane/mesh.py | 141 | 1 | 0 | live-endpoint | — | **B-** | CLAWX 24-phase signed state roots / edge verifier; enterprise path |
| 72 | 24 phases 14-17: signed state roots, an edge verifier with | clawx/control_plane/mesh.py | 141 | 1 | 0 | live-endpoint | — | **B-** | CLAWX mesh (duplicate listing); same evidence as #71 |
| 73 | Evolution 12 Byzantine Network Attack Lab | cain/trustnet/attacks.py | 477 | 18 | 1 | live-endpoint | — | **B-** | E12 Byzantine network attack lab; testing tool |
| 74 | Check, on your own machine, that a decision was really ord | sdk/python/src/cainstudio/verify.py | 430 | 0 | 0 | live-endpoint | — | **A-** | SDK/quorum offline verifier: public, zero-trust, 278 test references |
| 75 | E37 Byzantine Autonomous Federation -- core primitives | cain/govfederation/core.py | 414 | 2 | 0 | live-endpoint | — | **B-** | E37 Byzantine federation core; live-endpoint claim |
| 76 | adversarial soak on the LIVE multi-region cluster cain-mr- | scripts/cain42_bft/multi_region_soak.py | 397 | 1 | 1 | library | — | **B+** | multi-region adversarial soak harness against the live cluster |
| 77 | Trust Provenance Graph (Trust Kernel 3.0) | platform-gateway/cain_trust_provenance.py | 385 | 1 | 0 | library | — | **B-** | Trust Provenance Graph; wired |
| 78 | Epoch 6 -- World-State Integrity & Postcondition Reconcili | platform-gateway/cain_world_state_42.py | 339 | 2 | 0 | library | — | **B-** | Epoch 6 world-state integrity & reconciliation; wired |
| 79 | Evolution 7: predictive consequence governance (restrictio | cain45/l5/consequence_governance.py | 322 | 4 | 0 | library | — | **B-** | E7 predictive consequence governance (restriction-only) |
| 80 | Evolution 7: predictive consequence governance (restrictio | cain45/l5/consequence_governance.py | 322 | 4 | 0 | library | — | **B-** | E7 predictive consequence governance (duplicate); same as #79 |
| 81 | agent identity interop — SPIFFE-style workload identity, O | platform-gateway/cain_agent_identity.py | 294 | 1 | 0 | library | — | **B** | agent identity interop (SPIFFE/OAuth); wired |
| 82 | agent identity interop — SPIFFE-style workload identity, O | platform-gateway/cain_agent_identity.py | 294 | 1 | 0 | library | — | **B** | agent identity interop (duplicate listing); same evidence as #81 |
| 83 | Epoch 6: Action Proof Object & Proof-Carrying Agency | platform-gateway/cain_action_proof_epoch6.py | 294 | 1 | 0 | library | — | **B** | Epoch 6 Action Proof Object; wired |
| 84 | E40 hierarchical governance fabric (materialized form, for | cain/gov40/hierarchy.py | 292 | 1 | 0 | library | — | **C+** | E40 hierarchical governance fabric: materialised, in-memory populations only |
| 85 | Formal Safety Invariants Checker | platform-gateway/cain_formal_invariants_42.py | 286 | 1 | 0 | library | — | **B-** | formal safety invariants checker (42); wired |
| 86 | Formal Safety Invariants Checker | platform-gateway/cain_formal_invariants_42.py | 286 | 1 | 0 | library | — | **B-** | formal safety invariants checker (duplicate); same as #85 |
| 87 | Machine-Readable Trust Constitution | platform-gateway/cain_constitution_42.py | 280 | 2 | 0 | library | — | **B-** | machine-readable Trust Constitution; wired |
| 88 | E38 CLEAN-ROOM MISSION VERIFIER (ZERO CAIN imports; stdlib | platform-gateway/frontend/proof/bundle/e38-byzantine-mission-integrity-2026-10-05/cleanroom/verify_e38.py | 276 | 0 | 0 | live-endpoint | yes | **A-** | E38 clean-room mission verifier: zero CAIN imports; published bundle |
| 89 | Standalone verifier for a CAIN-42 Evolution 4 DAG cluster  | platform-gateway/frontend/proof/bundle/mcpgate-live-2026-09-27/verifiers/verify_dag_bundle.py | 242 | 1 | 0 | library | yes | **B+** | standalone PBFT-E4 DAG bundle verifier; published, no CAIN imports |
| 90 | Evolution 03 — canonical Byzantine Governance State, state | cain/govkernel/state.py | 238 | 1 | 0 | live-endpoint | — | **B-** | E3 canonical Byzantine governance state/state-machine |
| 91 | Evolution 20: Governed Agentic Civilization Fabric (machin | cain45/l5/institution_fabric.py | 3401 | 1 | 0 | library | — | **B-** | E20 governed agentic civilization fabric; 1 test import |
| 92 | Evolution 18: 4D Spatial Autonomy Fabric (predictive airsp | cain45/l5/spatial_autonomy.py | 3193 | 4 | 1 | library | — | **B-** | E18 4D spatial autonomy fabric; simulation scope |
| 93 | Evolution 21: Governed Open-Ended Intelligence Fabric (mac | cain45/l5/discovery_fabric.py | 2698 | 1 | 0 | library | — | **B-** | E21 open-ended intelligence fabric; research/discovery scope |
| 94 | Evolution 16: 4D Causal World Intelligence Fabric | cain45/l5/causal_world.py | 2042 | 3 | 1 | library | — | **B-** | E16 4D causal world intelligence; world-model/simulation scope |
| 95 | Evolution 29: Governed Agentic Internet Transaction Fabric | cain45/l5/transaction_fabric.py | 1767 | 1 | 0 | library | — | **B-** | E29 agentic internet transaction fabric |
| 96 | Evolution 29: Governed Agentic Internet Transaction Fabric | cain45/l5/transaction_fabric.py | 1767 | 1 | 0 | library | — | **B-** | E29 transaction fabric (duplicate listing); same as #95 |
| 97 | Evolution 36: the Machine Agency Exchange Fabric | cain45/l5/exchange_fabric.py | 1369 | 1 | 0 | library | — | **B-** | E36 machine agency exchange fabric |
| 98 | Evolution 36: the Machine Agency Exchange Fabric | cain45/l5/exchange_fabric.py | 1369 | 1 | 0 | library | — | **B-** | E36 exchange fabric (duplicate listing); same as #97 |
| 99 | Frontier 20 Common Runtime Substrate | cain/frontier/runtime.py | 1346 | 2 | 0 | live-endpoint | — | **A-** | Frontier-20 common runtime substrate: backbone of the 20 live products |
| 100 | Evolution 31: Universal Proof-of-Governance Fabric | cain45/l5/governance_proof.py | 1252 | 1 | 0 | library | — | **B-** | E31 universal proof-of-governance fabric; wired |

---

## 5. What the top tier actually is (and is not)

**A- — the deployable core (14).** Trust Runtime Kernel (#6), Governance Contract Fabric (#14), Agent Passport (#18), governance data plane (#20), Agent Hypervisor/ZoD (#21), PBFT engine (#29), CAG-L5 system governance (#31), the `/fabric/contracts/*` tenant API (#39/#40), capability supply-chain (#50), Agentic Internet Control Plane (#52), the SDK offline quorum verifier (#74), the E38 clean-room mission verifier (#88), and the Frontier-20 common runtime substrate (#99). These are genuinely enforced on the live gateway or published as verifiers that import none of CAIN's code. They are the reason the composite is not lower. **Even these cap at A-**: none has been reviewed by a third party, the deployment remains a single-provider cluster, and several live stages (`intent`, `approval`) are **off by default** per `/fabric/status`.

**B+/B — authoritative but not independently verified (26).** DAG consensus, primitive API, clean-room verifiers, the Authority Ledger, kernel self-defense, conformance/attestation engines and the identity interop layer. Real, tested, on real paths; missing published independent reproduction or operational hardening.

**B- — the research libraries (50).** The bulk of the `cain45/l5` Evolution fabric and the `cain/gov*` programs. Large (often 1 k–5 k LoC) and tested, but wired as internal libraries with single-digit test imports and no exposed product surface. These are **not products you can deploy**; they are the substrate the live pipeline draws on.

**C — thin / legacy / reference (8).** Legacy conformance spec, the OWASP-ASI invariant library, E6/E20-stage-1, the trifecta business pillar, cluster federation, E40 in-memory hierarchy, and the actuarial notary (#41: no insurer, no data feed).

**F — quarantined (2).** #37/#38 point at `quarantine/epoch14_concurrent_session_unverified/` — the directory name itself declares the session unverified. They must not be listed as products until re-verified.

---

## 6. Systemic blockers to "fully production-ready" (apply to most of the list)

1. **No third-party audit / reproduction.** Every verification artifact in the repo is same-project clean-room. Nothing is independently reproduced. (`C42-THIRD-PARTY-REVIEW` is `NOT_IMPLEMENTED`.)
2. **No enterprise compliance.** No SOC 2, no ISO 42001 certification; the repo says so itself.
3. **No live enforcement for the majority.** `/fabric/status` shows `intent` and `approval` disabled by default; only a subset of stages enforce.
4. **Single-cloud cluster.** "Multi-region" is four US regions on one provider — not independent failure domains.
5. **Pre-revenue, no customers, no filed patents.** Commercial readiness is unvalidated.
6. **Uniform placeholder metadata.** The homepage's own `status/tier/tests` fields are not real assessments and should be replaced with the measured statuses (as the "32 capabilities" section already does).

---

## 7. Reproduce

```bash
cd /root/cain-studio-platform
python3 scripts/audit/build_top100_readiness_grades.py          # rebuild grades JSON
python3 -m pytest tests/test_cain_trust_kernel_adversarial.py -q # 19 passed
curl -s https://cainstudio.online/fabric/status | head          # live stage truth
```

---

*Internal, evidence-based assessment. Not an appraisal, not a certification, not a guarantee. No third party has reviewed CAIN-42. Grades reflect the repository as inspected on 2026-10-07.*
