Related · tier-1 reading. For what a portable artifact actually is, see Portable Artifact.
H33-ZK-Phish uses zero-knowledge proofs to verify the authenticity of every link, email, and page you interact with — before you click. Not after.
ZK-Phish is zero-knowledge phishing protection — it proves a link, email, or page is authentic without exposing anything about the user or their browsing. It exists because the targeted phishing that slips past spam filters uses valid certificates and near-identical domains, so it has to be caught by cryptographic proof, not by heuristics that leak data. ZK-Phish is a supporting expression of the H33 ZK Platform, which proves without revealing.
The phishing emails that get through are the ones engineered to get through. They use valid certificates, real domains with one character changed, and pixel-perfect page clones.
Spam filters catch 95% of phishing emails. The other 5% — the targeted, well-crafted ones — get through. They use valid certificates, real domains with one character changed, and pixel-perfect page clones. Your employees click. Your data is gone.
Every link is verified before it opens. Every email is authenticated beyond DKIM/SPF. Every login page is compared against its real structure. The 5% that gets through spam filters gets caught by math.
Every check runs all 6 streams simultaneously. SHA3-256 commitments verify every claim. Zero-knowledge — nothing about the user or their data is revealed.
ZK-Phish intercepts, proves, and delivers a verdict — all before the page loads.
From simple typosquats to sophisticated multi-redirect chains. Every technique that bypasses spam filters — caught by proof.
Every stream completes before the browser finishes the TLS handshake. Cache hits from DashMap resolve in 80 nanoseconds.
Start free. Scale as your team grows. Every tier includes SHA3-256 zero-knowledge commitments.
| Tier | Volume | Price | Includes |
|---|---|---|---|
| Free | 1,000 checks/mo | $0 | URL + link verification |
| Pro | 50,000 checks/mo | $49/mo | All 6 proof streams + email |
| Business | 500,000 checks/mo | $299/mo | + page structure + QR + priority |
| Enterprise | Unlimited | Custom | + on-prem + custom commitments + SLA |
H33-ZK-Proven protects the connection. H33-ZK-Phish protects the content. Together, they verify both the pipe AND what flows through it. Man-in-the-middle attacks are stopped at the channel level. Phishing attacks are stopped at the content level. Nothing gets through.
Verify any URL, email, or page with a single request. Results in microseconds.
When to reach for ZK-Phish, where it stops, and what the examples on this page actually are.
Use it as a layer on top of your spam filters and email gateway, to reduce exposure to the targeted, well-crafted phishing that gets past them — valid certificates, one-character lookalike domains, pixel-perfect page clones. It is not a replacement for user training or for your existing email security, and no anti-phishing system can promise to catch every attack; ZK-Phish narrows the gap, it does not close it entirely. Do not use it when you need to compute directly over encrypted data instead of proving a claim about a link or page — use the FHE Platform instead.
No. ZK-Phish uses zero-knowledge proofs to verify that a link, email, or page is authentic without revealing user data. Computing over encrypted data is fully homomorphic encryption — the FHE Platform. ZK-Phish is a supporting expression of the ZK Platform, which proves without revealing; the portable, offline-verifiable evidence of each verdict is anchored by H33-74.
Illustrative. The domains, verdicts, scores, and reason codes shown on this page — including the chas3-secure.com example scoring 0.08 as PHISH — are sample outputs that show how the six proof streams score a request. They are not live results against a real target. A real verdict is produced only when you run an actual URL, email, or page through the API.
The proof streams verify claims about a link, page, or email using SHA3-256 commitments, without revealing anything about the user or their browsing behavior to the verifier. The math produces a verdict; it does not build a profile of the person.
1,000 free checks. No credit card.