Sample Report · Anonymized

See what an attacker already knows — before the strike.

Below is a real, authorized HYDRA engagement with the client fully anonymized — same depth, same discipline, zero PII. This is exactly the kind of report you receive when you work with us.

🛡️ Passive only🧾 Receipts-first🔒 GDPR · NDA-ready⏱ 2–3 days
SAMPLE · ANONYMIZED · ILLUSTRATIVE. This is a real HYDRA engagement with the client identity, domains, IP addresses and personnel fully redacted. The structure, methodology, lane logic and finding types are unchanged — this is exactly what a HYDRA report delivers. No live target is identified.
Tier 3 · Full Intelligence · Passive External Exposure
Sample Engagement Report
Authorized passive external exposure analysis — 5-lane HYDRA methodology, evidence-first.
HYDRAby TIA
Method: Passive OSINT
Lanes: 5
Findings: 12
Basis: real engagement, de-identified
Engagement · Scope

Client: [REDACTED] — a European small-to-mid business running several brands and a routine supplier-payment workflow. Sector, location, size, domains, IPs and people are generalized or redacted for this sample.

Authorization: written owner consent (assessment of the client's own public exposure). Method: 5-lane HYDRA — PASSIVE only (DNS/CT/DoH, public web, response headers, RDAP/whois, OSINT). No active scanning, no authentication attempts, no exploitation. Receipts-first, zero fabrication.

Declared up front: account-level breach confirmation and active characterization of the on-prem asset are not possible passively → scoped to a paid, signed Phase 2.

External Threat Exposure
ELEVATED
Four critical exposures on an otherwise healthy perimeter. The realistic money-loss path — email spoofing → fake supplier invoice — was confirmed independently across three lanes.
12
Findings
4
Critical / High
5
Recon Lanes
Passive
Method — 0 touch
Executive Summary — 4 real risks

The rest of the perimeter is reassuringly healthy (see the honest-negatives section). Four exposures carry real, exploitable risk:

  • 1On-prem remote-access box exposed to the internet. A management endpoint on consumer broadband, end-of-life web server, self-signed certificate — the single most attackable real asset. CRITICAL
  • 2Email authentication not enforced on both flagship domains (DMARC p=none). This opens a credible impersonation path — mail that can appear to come from leadership or a team address, on the real domain → Business Email Compromise / fake-invoice. Corroborated across infra + dark-int + people lanes. CRITICAL
  • 3Third-party look-alike domain with live mail. A recently-registered near-spelling of the brand, behind a privacy proxy, with active mail records — a ready-made phishing channel impersonating the client. CRITICAL
  • 4Shared role mailboxes; MFA posture not verifiable passively. A credential-hygiene pattern that, if unaddressed, makes risk #2 far easier to weaponize. HIGH
Findings by severity

Lanes — INFRA infrastructure · DARK-INT breach & dark-web · WEB web surface · PEOPLE human/OSINT · BRAND domains & impersonation.  SeverityCRITICAL / HIGH exploitable now · MEDIUM hardening.

F1On-prem remote-access endpoint exposed
[ INFRA ]
A deliberate DNS record points a management hostname at consumer broadband (not a datacenter). Passive fingerprint: end-of-life web server, self-signed certificate (multi-year). Consistent with an on-prem management / VPN / back-office box — internet-reachable and unadvertised in certificate transparency; its internal network adjacency is inferred, not passively confirmed. → identify it, place behind a zero-trust broker, patch/replace, pull off the public net. What it actually runs = Phase 2 (signed scope).
F2DMARC p=none on both flagship domains → BEC
[ INFRA · DARK-INT · PEOPLE ]
Both flagship domains publish p=none (monitor, no enforcement) despite valid SPF and DKIM. Mail can be sent as From: <anyone>@[client-domain] with no cryptographic rejection — a credible impersonation path (final delivery still depends on the recipient's filtering, but the sender's own domain raises no barrier). For any business that pays suppliers, this is a fake-invoice / payment-redirect vector. → add aggregate reporting, confirm sender alignment, then move none → quarantine → reject on both.
F3Look-alike domain, mail-capable
[ BRAND · INFRA · WEB ]
A near-spelling of the brand, registered ~6 months ago behind a US privacy proxy, on a different hosting stack than the client — every fingerprint diverges, so it is not theirs. Storefront disabled, but live MX records present: a disabled store ≠ disabled mail = a ready phishing/BEC channel. → attempt acquisition; else abuse-report + registry dispute; monitor records; warn staff and customers.
F4Shared role mailboxes + MFA hygiene
[ DARK-INT · PEOPLE ]
Multiple shared role mailboxes are visible (generic function addresses such as info@ / orders@). These tend to be the highest-reuse, lowest-rotation accounts — shared passwords, reused as portal logins; per-user MFA cannot be confirmed passively and is a Phase-2 check. One leaked shared password would let an attacker read supplier correspondence → invoice fraud (amplifies F2). → enforce org-wide 2-step verification; migrate shared → owner+delegation; rotate now.
F5Expired / wrong-owner TLS + HTTPS→HTTP downgrade
[ WEB · INFRA ]
Two secondary sites present an expired certificate for an unrelated owner and downgrade https → plaintext http. Erodes trust and trains users to click through certificate warnings. → valid certs or TLS-valid edge + clean redirects.
F6Shop domain with no SPF / no DMARC
[ INFRA ]
A secondary shop domain is fully spoofable (no SPF, no DMARC) — a clean channel for customer-facing phishing. → v=spf1 -all + p=reject.
F7Single publicly-named finance identity
[ PEOPLE ]
Public professional profiles name exactly one finance decision-maker — both an impersonation target and a spoofing vector for payment fraud. → authenticate finance comms; callback/codeword for any payment or bank-detail change; reduce public identifiability.
F8Third-party processor cascade
[ DARK-INT ]
Several third-party processors sit in the CRM / payments / mail path (customer PII, payment context). A breach at any one leaks client data indirectly; password reuse turns it into a mail-tenant compromise. → inventory staff↔processor logins; unique passwords + MFA; breach-monitor admin accounts.
F9Content-Security-Policy report-only; weak Referrer-Policy
[ WEB · INFRA ]
The primary site ships CSP in report-only mode and a permissive Referrer-Policy. → promote CSP to enforced (the sister site already does); tighten Referrer-Policy.
F10Defensive-registration & listing gaps
[ BRAND ]
Several obvious brand variants (.com/.eu/.net + alternate spellings) are unregistered and a social handle is unclaimed — open ground for impersonation. → register defensives, claim the handle, verify business listings.
F11Broken backup MX record
[ INFRA ]
A secondary MX points to a non-resolving host (possibly tied to the F1 on-prem box). → fix or remove.
F12Account-level breach exposure — UNVERIFIED
[ DARK-INT ]
Free-tier sources show no public hit (confirmed negative via dorks + public leak lists), but regional combolists sit behind paid databases. This is absence of public evidence, not evidence of absence. → Phase 2 paid pass (breach DBs, Split-Brain on any hit).
Honest negatives — what's actually solid

A report that is all-red is selling fear, not truth. Here's the genuinely good news, evidence-backed:

  • Managed SaaS platforms eliminate whole classes of self-hosted vulnerability; card data sits with the payment provider.
  • No secrets, stack traces or sensitive paths exposed in page source or robots/sitemap.
  • Mail baseline is healthy — SPF + DKIM present on owned domains; two sub-brands already enforce DMARC.
  • No social-media impersonation found; aggregator listings consistent.
  • The owned domain portfolio is genuinely owned and correctly configured.
Remediation plan — severity-ordered
🔴 Now — quick wins
  1. Investigate + isolate the on-prem remote box (zero-trust, patch, off the public net).
  2. DMARC none → quarantine → reject on both flagship domains (after alignment monitoring).
  3. Decide the look-alike domain: acquire, or abuse-report + dispute; warn staff and customers.
  4. Enforce org-wide 2-step verification; rotate shared mailbox passwords.
  5. Callback-verification + two-person rule for ANY payment or bank-detail change — the single highest-leverage human control, and it defends even against look-alike-domain attacks that DMARC can't stop.
🟡 Soon
  1. Fix expired certs / HTTPS downgrade; add SPF + DMARC to the shop domain.
  2. Enforce CSP + tighten Referrer-Policy; register defensive domains + claim listings/handle.
  3. Inventory staff↔processor logins; unique passwords + MFA everywhere.
🟢 Hygiene
  1. Remove the broken backup MX; consolidate the domain portfolio under one registrar with 2FA + auto-renew.
  2. Verify DNSSEC on remaining domains; confirm consent-gating on tracking.
Phase 2 — where this goes next (paid, signed scope)
  • Active vulnerability assessment of the on-prem box — what it runs, harden or remove.
  • Paid breach-database pass — confirm or deny account-level credential exposure.
  • DMARC enforcement rollout + staff BEC-awareness training ("finance never changes a bank account by message").
  • Ongoing monitoring — DMARC aggregate reports + brand/domain monitoring.
Methodology & discipline

Receipts-first, zero fabrication. Every 🔴/🟡 is backed by a passive DNS record, RDAP entry, response header or certificate read. Honest negatives are stated as negatives. Explicitly-unverified items (what the on-prem box runs; account-level breach; true owner of the look-alike domain) are flagged as unverified and routed to a signed Phase 2 — never guessed, never inflated.

Passive only. No active exploitation was attempted. All findings derive from publicly available information and open-source intelligence.

HYDRA cuts heads of dragons — not merchants
We find what an attacker already knows about you, prove it with evidence, and hand you a closed-loop path to fix it. The first directional signal is free; full proof and remediation are scoped.

Want to see your own?

We run this same 5-lane passive assessment on your domains. The first directional signal is free — full proof and remediation are scoped.