Midnight Proofsproof Generation

Proof Generation Flow

1. Receive Request - Accept witness data and circuit identifier
2. Load Circuit - Retrieve proving key for the circuit
3. Generate Proof - Execute the prover with witness inputs
4. Return Proof - Send proof bytes back to client

// Server-side proof generation overview
async function generateProof(request: ProofRequest): Promise<ProofResponse> {
  const { circuitId, witness } = request;

  // Load circuit proving key
  const provingKey = await loadProvingKey(circuitId);

  // Generate the proof (CPU intensive)
  const proof = await prover.prove(provingKey, witness);

  return { proof, circuitId };
}

Witness Data Handling

Witness data contains private inputs to the circuit. When handling server-side:

• Encrypt in transit - Use TLS and consider additional encryption
• Minimize retention - Delete witness data immediately after proof generation
• Audit logging - Log proof requests (not witness content) for debugging
• Access control - Authenticate and authorize proof requests

References

Examples

Quick Start

1. Set Up Prover Service

import { createProver } from '@midnight-ntwrk/midnight-js-prover';

const prover = await createProver({
  // Path to circuit proving keys
  circuitKeysPath: './circuit-keys',
  // Memory limit for proof generation
  memoryLimit: 4096, // MB
});

2. Create Proof Endpoint

import express from 'express';

const app = express();
app.use(express.json());

app.post('/api/prove', async (req, res) => {
  const { circuitId, witness } = req.body;

  try {
    const proof = await prover.prove(circuitId, witness);
    res.json({ success: true, proof: proof.toString('base64') });
  } catch (error) {
    res.status(500).json({
      success: false,
      error: error.message
    });
  }
});

3. Handle Async Requests

// For long-running proofs, use async pattern
app.post('/api/prove/async', async (req, res) => {
  const { circuitId, witness } = req.body;

  // Create job ID
  const jobId = crypto.randomUUID();

  // Queue proof generation
  await proofQueue.add({ jobId, circuitId, witness });

  // Return immediately with job ID
  res.json({ jobId, status: 'queued' });
});

// Poll for status
app.get('/api/prove/status/:jobId', async (req, res) => {
  const job = await proofQueue.getJob(req.params.jobId);

  if (!job) {
    return res.status(404).json({ error: 'Job not found' });
  }

  if (job.status === 'completed') {
    return res.json({ status: 'completed', proof: job.result });
  }

  res.json({ status: job.status });
});

Common Patterns

Request Validation

import { z } from 'zod';

const ProofRequestSchema = z.object({
  circuitId: z.string().min(1),
  witness: z.record(z.unknown()),
  timeout: z.number().optional().default(60000),
});

function validateProofRequest(body: unknown) {
  return ProofRequestSchema.parse(body);
}

Timeout Handling

async function proveWithTimeout(
  circuitId: string,
  witness: WitnessData,
  timeoutMs = 60000
): Promise<Proof> {
  const controller = new AbortController();
  const timeout = setTimeout(() => controller.abort(), timeoutMs);

  try {
    return await prover.prove(circuitId, witness, {
      signal: controller.signal,
    });
  } finally {
    clearTimeout(timeout);
  }
}

Health Check Endpoint

app.get('/health', async (req, res) => {
  try {
    // Verify prover is ready
    const isReady = await prover.isReady();

    res.json({
      status: isReady ? 'healthy' : 'degraded',
      proverReady: isReady,
      uptime: process.uptime(),
    });
  } catch (error) {
    res.status(503).json({ status: 'unhealthy', error: error.message });
  }
});

Error Categories

enum ProofErrorCode {
  INVALID_WITNESS = 'INVALID_WITNESS',
  CIRCUIT_NOT_FOUND = 'CIRCUIT_NOT_FOUND',
  PROOF_GENERATION_FAILED = 'PROOF_GENERATION_FAILED',
  TIMEOUT = 'TIMEOUT',
  OUT_OF_MEMORY = 'OUT_OF_MEMORY',
}

class ProofError extends Error {
  constructor(
    public code: ProofErrorCode,
    message: string,
    public details?: unknown
  ) {
    super(message);
    this.name = 'ProofError';
  }
}

Performance Considerations

Related Skills

• proof-verification - Verify generated proofs
• proof-caching - Cache proof components for performance
• prover-optimization - Tune prover for production workloads

Related Commands

None currently defined.