Founder built

TrojanStrike

Permissioned attack simulation — attack yourself first. The marketing site is live; the strike engine is proprietary software in active development.

trojanstrike.com
TrojanStrike homepage — permissioned attack simulation headline over a live strike terminal
SectorWeb security — attack simulation
StackNext.js
Live URLtrojanstrike.com

What we built — and where it's at

TrojanStrike is the marketing site for a proprietary attack-simulation engine currently in development. The product vision: prove domain ownership, enter one hostname, and a permissioned, rate-limited, read-only strike runs 56 checks across six escalation classes — with every request logged and streamed live. It renders one of four verdicts (INSIDE_THE_WIRE, BREACHED, AT_THE_DOOR, HELD) backed by the proving request, the impact and a concrete fix with a repro cURL. Recon is always free — and so is your first strike.

Verdicts, not scores

Every strike will end in one of four verdicts — no scores, no hedging. The build is Next.js with a stark, evidence-first design language: white ground, near-black display type, and a single violet accent reserved for status language. The homepage stages the strike as a scrolling four-phase narrative with a live lattice animation, and the Capabilities page commits to proof over promises — publishing benchmark scores and the misses. With the engine still in development, the site does the selling with the record it has already put on the table.

Details that matter

0 requests before attestation. '0 requests before attestation' — ownership is verified before any probe is sent. Recon is free, and so is your first strike.

Four verdicts, no scores. The verdict taxonomy is explicit four-state language: INSIDE_THE_WIRE, BREACHED, AT_THE_DOOR and HELD — one verdict per engagement, 'No scores, no hedging.'

Proof with every finding. Every finding ships with proof: the verdict, the proving request, the impact and the fix — including a repro cURL. The sample BOLA/IDOR finding (CWE-639) is published.

Benchmarks on record. The Capabilities page publishes benchmark scores on the record: OWASP BenchmarkJava 2,740 cases with 0 false positives, 26/26 Juice Shop and 16/16 WebGoat verdicts — misses published by name.

Four-phase strike narrative. A scrolling 'geometry of a strike' narrative walks through four phases — RECON, PROBE, STRIKE and VERDICT.

Live lattice hero. The hero renders a live status panel — 'CHECKS 56 · REQUESTS 4.6K · VERDICT PENDING' — over an animated lattice of nodes.

Published sample report. A full sample report is linked from the homepage at /sample-report, with /login and /strike routes behind the marketing page.

The proof ladder. The Capabilities page climbs a proof ladder — ground-truth benchmark, live owned targets, autonomous red-team grading — with false-alarm counts at each step.

Stack evidence: Next.js app — __next markers, immutable chunks, /_next/image device-picture-loader params.

How a strike works

Four steps. No black box.

01

Verify ownership

Prove the domain is yours. No attestation, no packets.

02

Enter target

One hostname — that's the whole setup.

03

The strike runs

56 attack & recon checks fire, rate-limited and read-only. The engine fingerprints your stack mid-scan and re-orders the surface to fit it.

04

Verdict + fix

Every strike ends in one of four verdicts — INSIDE_THE_WIRE, BREACHED, AT_THE_DOOR or HELD — with the evidence and the fix.

TrojanStrike homepage — permissioned attack simulation headline over a live strike terminal

One client per industry, per area

Want this for your business?

Every build starts the same way: you describe it, we quote it fixed, and you see the finished design before you pay a cent.

Answered personally, same business day.