Claude Code Plugins

Community-maintained marketplace

Feedback

>-

Install Skill

1Download skill
2Enable skills in Claude

Open claude.ai/settings/capabilities and find the "Skills" section

3Upload to Claude

Click "Upload skill" and select the downloaded ZIP file

Note: Please verify skill by going through its instructions before using it.

SKILL.md

name privacy-patterns
description Implement privacy-preserving patterns using zero-knowledge proofs on Midnight Network. Use when designing private data handling, commitment schemes, nullifiers, or selective disclosure. Triggers on privacy, ZK proof, commitment, nullifier, or confidential computing questions.

Privacy Patterns for Midnight

Design and implement privacy-preserving applications using zero-knowledge proofs.

Core Privacy Model

Concept Description Visibility
Public Ledger state Everyone
Private Circuit inputs Only prover
Witness Prover-provided data Only prover
Disclosed Explicitly revealed Everyone

Reference Files

Topic Resource
Zero-Knowledge Basics references/zk-fundamentals.md
Commitment Schemes references/commitments.md
Nullifier Patterns references/nullifiers.md
Selective Disclosure references/selective-disclosure.md

Pattern Overview

1. Commitment Scheme

Hide a value while binding to it:

commitment = persistentCommit(value, randomness);
// Later: prove you know the opening

2. Nullifier Pattern

Prevent double-use without revealing identity:

nullifier = transientHash(secret, commitment);
// Can only be generated once per commitment

3. Selective Disclosure

Prove properties without revealing data:

// Prove over 18 without revealing actual age
disclose(age >= 18);  // Only boolean is public

4. Merkle Membership

Prove membership in a set without revealing position:

// Verify path from leaf to root
assert verifyMerklePath(leaf, proof, root);

Quick Examples

Private Balance Check

// Only reveal if balance is sufficient, not actual amount
export circuit checkFunds(balance: Uint<64>, required: Uint<64>): Boolean {
  return disclose(balance >= required);
}

Anonymous Voting

export circuit vote(voter: Bytes<32>, choice: Boolean): [] {
  // Voter identity disclosed (prevents double voting)
  hasVoted.insert(voter);
  // Choice remains private, only totals change
  if (choice) { yesCount = yesCount + 1; }
}

Commitment-Reveal

witness randomness: Field;

// Phase 1: Commit
export circuit commit(value: Uint<64>): Field {
  return persistentCommit(value, randomness);
}

// Phase 2: Reveal
export circuit reveal(value: Uint<64>, commitment: Field): [] {
  assert persistentCommit(value, randomness) == commitment;
  disclose(value);
}

Privacy Best Practices

  • ✅ Use witness for data that should never appear on-chain
  • ✅ Use persistentCommit (with randomness) to hide values
  • ✅ Use nullifiers to prevent double-actions
  • ✅ Disclose only what's necessary (prefer booleans)
  • ❌ Don't store unhashed sensitive data on ledger
  • ❌ Don't use predictable randomness in commitments
  • ❌ Don't reveal intermediate values unnecessarily

Privacy Levels

┌────────────────────────────────────────────────┐
│ Level 0: Fully Public                          │
│ - All data visible on-chain                    │
├────────────────────────────────────────────────┤
│ Level 1: Hidden Values                         │
│ - Commitments on-chain, values private         │
├────────────────────────────────────────────────┤
│ Level 2: Unlinkable Actions                    │
│ - Nullifiers prevent linking actions           │
├────────────────────────────────────────────────┤
│ Level 3: Anonymous Membership                  │
│ - Merkle proofs hide set position              │
└────────────────────────────────────────────────┘

When to Use Each Pattern

Pattern Use Case
Commitment Sealed bids, hidden votes before reveal
Nullifier Preventing double-spend, one-time tokens
Merkle Proof Membership in allowlist without revealing identity
Selective Disclosure Age verification, credential proofs

Resources