Preview

# Demo: billing-gateway — post-quantum migration plan

> Generated by PQC Triage 1.0.0 · engine `hndl@1.0.0` · 2026-10-04

This plan is an engineering aid, not an assurance. Every number below is produced by a deterministic function you can read and re-run; none of it is a certification, and it does not tell you whether a specific system is safe.

## Deadline arithmetic

Mosca's rule: if data must stay confidential for **X** years after capture, the migration takes **Y** years, and a cryptographically relevant quantum computer arrives in **H**, work must start by **H − (X + Y)**.

| Setting | Value | Source |
| --- | --- | --- |
| Quantum-capability horizon (H) | 2033 | Analyst assumption, editable in Settings |
| NIST deprecates 112-bit public key | 2030 | [NIST IR 8547](https://csrc.nist.gov/pubs/ir/8547/ipd) |
| NIST disallows quantum-vulnerable public key | 2035 | [NIST IR 8547](https://csrc.nist.gov/pubs/ir/8547/ipd) |
| NIST disallows 112-bit-level symmetric | 2030 | [NIST IR 8547](https://csrc.nist.gov/pubs/ir/8547/ipd) |

## Portfolio

- **Weighted exposure score:** 58 / 100 (high)
- **After the recorded decisions:** 58 / 100 (high)
- **Surfaces:** 12, of which 0 triaged and 6 already past their start date
- **First thing that breaks:** in -3.8 years
- **Highest risk:** Math.random() for a secret in source at 90

## Surfaces

| Surface | Primitive | Role | X (yrs) | Score | Start by | Decryptable from | Replacement | Decision |
| --- | --- | --- | --- | --- | --- | --- | --- | --- |
| ECDH P-256 in source | ecdh-p256 | session-establishment | 5 | 77 | 2025 | 2028 | ML-KEM-768 (FIPS 203), hybrid with X25519 | untriaged |
| RSA-2048 in source | rsa-2048 | session-establishment | 5 | 78 | 2025 | 2028 | ML-KEM-768 (FIPS 203), hybrid with X25519 | untriaged |
| 3DES in source | triple-des | data-at-rest | 7 | 78 | 2025 | 2026 | AES-256-GCM | untriaged |
| MD5 in source | md5 | unknown | 5 | 86 | 2027 | 2028 | SHA-256 | untriaged |
| Math.random() for a secret in source | math-random | key-generation | 5 | 90 | 2028 | 2028 | crypto.randomBytes / secrets.token_bytes | untriaged |
| jsonwebtoken@9.0.2 · HMAC-SHA-256 | hmac-sha256 | session-establishment | 5 | 14 | 2028 | 2028 | Keep HMAC-SHA-256 | untriaged |
| node-forge@1.3.1 · RSA-2048 | rsa-2048 | session-establishment | 5 | 79 | 2025 | 2028 | ML-KEM-768 (FIPS 203), hybrid with X25519 | untriaged |
| bcrypt@5.1.1 · bcrypt | bcrypt | password-storage | 0 | 4 | 2033 | 2033 | Keep bcrypt or move to Argon2id | untriaged |
| elliptic@6.5.4 · ECDSA P-256 | ecdsa-p256 | code-signing | 10 | 81 | 2020 | 2023 | ML-DSA-65 (FIPS 204) | untriaged |
| libsodium-wrappers@0.7.15 · ChaCha20-Poly1305 | chacha20-poly1305 | data-in-transit | 3 | 6 | 2030 | 2030 | Keep ChaCha20-Poly1305 | untriaged |
| tweetnacl@1.0.3 · X25519 | x25519 | session-establishment | 5 | 76 | 2026 | 2028 | ML-KEM-768 (FIPS 203), hybrid with X25519 | untriaged |
| HMAC-SHA-256 in source | hmac-sha256 | session-establishment | 5 | 14 | 2028 | 2028 | Keep HMAC-SHA-256 | untriaged |

## Why each score is what it is

### ECDH P-256 in source

ECDH P-256 is 1.8 years past the point where work should have started; start migration by 2025 is already in the past.

Quantum strength: **0 bits**. Häner et al., Estimating the resource requirements of breaking ECC, PQCrypto 2020 / IACR ePrint 2020/1193.

| Factor | Weight | Severity | Points | Why |
| --- | --- | --- | --- | --- |
| Harvest-now-decrypt-later window | 0.26 | 0.95 | 24.7 | Data protected for 5 years is decryptable from 2028 at a 2033 quantum-capability horizon, so migration must start by 2025. _X=5, Y=3, H=2033; NIST disallows this strength level in 2035._ |
| Quantum security strength | 0.22 | 1.00 | 22.0 | ECDH P-256 is broken outright by Shor's algorithm; there is no quantum security level left to preserve. _NIST SP 800-57 Part 1 Rev. 5, Key Management; symmetric strength halves under Grover._ |
| Migration headroom | 0.18 | 0.88 | 15.8 | 1.2 years of headroom r

… truncated in the preview; the download is complete.