Security

Curated in-app view

Source file linked

Document route

Cryptographic Confidence Engine

Deterministic scoring model for how ZK, REFHE, MagicBlock, Fast RPC, and hardening surfaces change the confidence profile of each proposal pattern.

Document context

Interpretation layer only; it does not replace audit, custody, or formal security proofs.

Audience: Judges, security reviewers, operators, buyers

Try in product Open verification route Open generated proof

Open raw file

PrivateDAO Cryptographic Confidence Engine

This document defines the product-facing scoring model that explains why one PrivateDAO proposal pattern is stronger, more private, or more reviewer-complete than another.

It is intentionally truth-aligned.

It is not:

a claim of impossible-to-break security
a formal proof that the protocol cannot fail
a substitute for external audit closure or production custody evidence

It is a deterministic scoring layer that combines the repository's real surfaces:

ZK
REFHE
MagicBlock
Fast RPC
Governance Hardening V3
Settlement Hardening V3
live proof packets
audit and launch-boundary surfaces

Why it exists

PrivateDAO now exposes many strong primitives, but they do not all contribute in the same way for every proposal.

Examples:

a confidential payroll flow benefits heavily from REFHE
a gaming rewards corridor benefits heavily from MagicBlock
a grant committee flow benefits more from ZK and reviewer proof surfaces than from encrypted payout envelopes

The engine makes those differences visible instead of pretending that every proposal has the same cryptographic posture.

The four dimensions

The engine scores four dimensions and combines them into a single 0-100 score.

Dimension	Weight	Meaning
Privacy depth	`28%`	How strongly the proposal path reduces intent leakage and protects private signal collection
Enforcement depth	`28%`	How strongly policy snapshots, settlement rules, and proof anchors constrain the path
Execution integrity	`24%`	How strongly runtime evidence and bounded execution surfaces support correct execution
Reviewer confidence	`20%`	How easy it is for a reviewer to inspect proof, trust, and launch-boundary evidence

The factor model

Privacy depth

Commit-reveal voting
ZK review overlay
REFHE confidential envelope
Proposal-bound proof anchors

Enforcement depth

Governance Hardening V3
Settlement Hardening V3
Proposal-bound proof anchors
MagicBlock settlement evidence
REFHE execution boundary

Execution integrity

Fast RPC indexed runtime
MagicBlock corridor evidence
REFHE envelope readiness
Baseline live proof
Dedicated V3 proof

Reviewer confidence

Baseline live proof
Dedicated V3 proof
Audit packet
Launch boundary remains explicit

Formula

For each dimension:

Add the weights of the active factors
Divide by the total factor weight for that dimension
Convert that ratio to a 0-100 dimension score

Then compute the total:

total =
  privacyDepth * 0.28 +
  enforcementDepth * 0.28 +
  executionIntegrity * 0.24 +
  reviewerConfidence * 0.20

The result is rounded to the nearest integer.

Interactive policy composer

The Next.js security surface includes an interactive composer that lets the operator or reviewer toggle:

Commit-reveal
ZK review
Proposal-bound proof anchors
Governance V3
Settlement V3
REFHE
MagicBlock
Fast RPC
Live proof
V3 proof
Audit packet
Explicit launch boundary

This does not create any new protocol capability by itself.

It exists to answer a practical product question:

> if we enable or disable one of these rails, how much confidence do we lose, and in which dimension?

It also includes product-pack presets for:

Grant Committee
Fund Governance
Gaming DAO
Enterprise DAO

These presets are not magic templates. They are opinionated starting points for the composer, and the operator can still override each rail manually.

Score bands

Band	Range	Meaning
`Advanced`	`80-100`	Strong multi-layer posture with explicit proof and execution boundaries
`Strong`	`62-79`	Good posture with meaningful hardening, but some dimensions remain lighter
`Foundational`	`<62`	Useful surface, but not enough depth to market as a high-confidence cryptographic path

Scenario reading

Confidential payroll

High score because it combines:

private voting
ZK review overlay
REFHE envelope
Governance V3
Settlement V3
Fast RPC
baseline + V3 proof

Private grant committee

High reviewer confidence and privacy, but lower execution-integrity contribution from:

no REFHE payout envelope
no MagicBlock corridor evidence

Gaming rewards corridor

Strong execution-integrity and settlement semantics from:

MagicBlock
Settlement V3
Fast RPC

But lower privacy depth because this path usually does not use:

REFHE
ZK review overlay

What the engine does not claim

The engine does not convert the project into any of the following if they are not already true:

full on-chain verifier CPI
anonymous private treasury execution
hidden tally replacement
mainnet custody completion
audit sign-off completion
impossible-to-exploit protocol behavior

Reviewer shortcut

Use this engine together with:

The engine is useful because it turns those surfaces into a single, explainable reading model instead of leaving them as disconnected documents.

Related next docs

Agentic Treasury Micropayment Rail

Operational brief for DAO-controlled micropayment batches, showing how approved policy becomes batched stablecoin settlement with judge-visible runtime proof and telemetry continuity.

Reviewer Fast Path

Shortest reviewer path across live proof, V3 hardening, trust links, and launch boundary surfaces.

Reviewer Telemetry Packet

Generated reviewer-visible route into telemetry, hosted reads, runtime evidence, indexed governance, and the infrastructure value layer behind PrivateDAO.