Proof & Resources

Evaluation Library & Technical Resources

Methodology, evidence classification, and reproducible verification records. No invented benchmarks, metrics, or market claims.

Available

All evaluation records on this page correspond to verified, deterministic test outputs and repository gate artifacts.

Verification & Evaluation Records

Web availabledeterministic-gate

Mission Breadth Forensic Reconstruction Verification

Cross-validated verification of recovered mission-breadth data schemas, catalog parity, and visual layout rendering.

Methodology

Automated SHA-256 byte-for-byte artifact comparison, full TypeScript schema compilation, and 60-case Playwright visual QA.

Reproducibility Command

npm run test:unit && npm run test:integration && npm run test:coverage

Limitations

Evaluates local web platform rendering and schema validity only.
Does not evaluate production edge deployment or live cloud agent execution.

Web availableverification-ledger

Script Budget & Framework Floor Alignment

Empirical measurement of shared JavaScript bundle weight against Next.js framework baseline and route budgets.

Methodology

Automated Playwright performance harness running on production standalone build server with simulated 4G throttling.

Reproducibility Command

npm run test:performance && npm run test:performance:ledger

Limitations

Script budget is 100 kB for home and trust surfaces based on ADR-0001.
Framework floor is measured at 86.9 kB for shared Next.js runtime chunks.

Web availableaudit-log

Claims Registry Schema & Approval Boundary Audit

Automated linting and validation of public statements against registered risk levels and human approval requirements.

Methodology

Fail-closed AST parsing of public web surfaces matched against specs/content/claims.registry.json.

Reproducibility Command

npm run validate:claims

Limitations

Identifies 4 strategic/performance claims blocked pending recorded human legal/security approvals.
Does not auto-grant approvals without verified operator signatures.

A proof-first technical evaluation path

Start with local records and their limitations, then inspect architecture, protocols, security responsibilities, and evidence concepts. These pages describe the model; they do not provide an SDK, live OS connection, customer deployment proof, or data room.

Architecture and boundaries Protocol maturity Request a technical briefing

Technical Resources & Architecture Specs

Omnacta Technical Architecture & Outcome Authority Briefing

Detailed architectural breakdown of Kernel authority, decision capsules, and evidence-bound mission execution.

Target Audience: Engineering Leaders & System Architects

Source Ref: docs/07-technical-architecture.md

Access Resource →

Evidence Model & Claim Verification Specification

Governing specification for evidence generation, claims validation, and fail-closed publication gates.

Target Audience: Security, Legal & Compliance Evaluators

Source Ref: docs/06-content-evidence-governance.md

Access Resource →

GlassBox Causal Observability & Control Positioning

Planned positioning for GlassBox causal observability sharing Kernel authority and the mission ledger.

Target Audience: Enterprise Evaluators

Source Ref: docs/14-glassbox.md

Access Resource →

Controlled Enterprise Diligence Specification

Typed specification governing identity, authorization, document custody, watermarking, and audit retention.

Target Audience: Enterprise Security & Legal Teams

Source Ref: data/proof-diligence/diligence.ts

Access Resource →