Download the sample report files: Sample Penetration Testing Report (.pdf)

Enterprise customer security teams, SOC 2 Type II auditors, and PCI DSS Qualified Security Assessors read enough penetration test reports that they can spot an automated scanner export in less than a minute. A sixty-page PDF padded with informational HTTP header notes and unverified server version banners can look thorough from across the room. Yet none of those pages tell you whether an authenticated user in Tenant B can export Tenant A's bank routing numbers.

When you ask a prospective testing firm for a sample deliverable, what should you actually look for? Maybe the practical test comes down to two questions: can one of your backend developers replay every finding locally in five minutes without booking a follow-up call, and can you hand proof of testing to an enterprise buyer without exposing internal API routes? Pair our guide on scoping a web application penetration test with the synthetic walkthrough below before you sign a statement of work. (Note: acmeledger.example, Org_Alpha, and all commit hashes and payloads below are synthetic examples written by briskData engineers to show report structure, never customer data.)

Manual Pentest Deliverable vs. Automated Scanner Export

Requesting a redacted sample report upfront saves unpleasant surprises at the end of an audit window. Five details separate manual engineering work from a dressed-up scan:

Report Element Manual Application Pentest Report Unauthenticated Scanner Export
Scope and build traceability Records exact target hostnames, test window, tester egress IPs, multi-tenant role accounts, and Git commit SHA Lists a single root domain scanned unauthenticated for an hour
Authorization and workflow depth Tests cross-tenant BOLA/IDOR, vertical role escalation, webhook spoofing, and state-machine race conditions Has no concept of tenants, roles, billing state, or domain rules
Reproduction evidence Redacted raw HTTP request/response pairs and exact steps any backend developer can replay in five minutes Generic banner warnings ("server version may be affected") mixed with false positives
Remediation specificity Concrete code-level diffs matched to the target stack (Python/FastAPI, TypeScript, Rails, .NET, PostgreSQL) Boilerplate encyclopedia blurbs advising teams to "validate all user input"
Procurement shareability Includes dated Retest Verification status plus a standalone two-page Letter of Attestation for prospects Leaves open findings unresolved or forces you to expose internal URLs to prospective buyers

Executive Summary and Reproducible Scope Block

Executives, cyber-insurance underwriters, and procurement reviewers rarely read past the first three pages unless something looks off. A clear Executive Summary states plainly what was tested, describes the business risk without dumping raw stack traces into prose, and tabulates findings by severity right next to their post-retest status.

Directly underneath, the Scope and Rules of Engagement block pins the test to exact technical coordinates so engineers know what build was running. In our synthetic report, testing targeted https://staging-app.acmeledger.example and https://staging-api.acmeledger.example on release v2.18.4 (Git commit 4d91b02e) during October 12 through 19, 2026. The engagement followed a manual grey-box methodology aligned with OWASP WSTG, OWASP ASVS v4.0 Level 2, and the OWASP API Security Top 10 (2023), using two isolated synthetic tenants (Org_Alpha_ID: 1042 and Org_Bravo_ID: 1043) provisioned across Owner, Billing Manager, Read-Only Auditor, and scoped API-token credentials.

Sample Technical Findings With HTTP Proof and Code Diffs

Every finding in the downloadable PDF pairs CVSS v3.1 vector context and CWE/OWASP mappings with affected routes, reproducible HTTP requests, and stack-specific remediation diffs. Here are two abridged findings from the synthetic acmeledger.example deliverable.

Sample Finding BD-2026-01: Cross-Tenant Invoice Export via Unscoped Object UUID

  • Severity: Critical (CVSS v3.1 9.1, CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:C/C:H/I:H/A:N) · Class: CWE-639 / OWASP API1:2023 (BOLA)
  • Affected endpoints: GET /api/v2/invoices/{invoice_uuid}/export and PATCH /api/v2/invoices/{invoice_uuid}
  • Retest status: Remediated and verified on October 27, 2026 against commit 91c4e80b

It is not unheard of for an application's main list views to filter cleanly by tenant while a secondary export or update handler forgets the tenant clause. In this sample case, dashboard list queries filtered records by organization_id, yet the JSON/PDF export and update handlers looked up invoice_uuid by primary key alone after checking only that the caller held billing:read inside their own organization. Whenever an invoice UUID surfaced in a shared receipt link or webhook log, any authenticated user in Org_Bravo could read or modify vendor ACH details belonging to Org_Alpha:

# Authenticated as Read-Only Auditor in Org_Bravo (org_id: 1043)
# Requesting an invoice UUID belonging to Org_Alpha (org_id: 1042)
GET /api/v2/invoices/7f9c2d10-4b88-41e0-9a12-08d4c1109e33/export HTTP/1.1
Host: staging-api.acmeledger.example
Authorization: Bearer eyJhbGciOiJSUzI1NiIsInR5cCI6IkpXVCJ9...[REDACTED_BRAVO_JWT]
Accept: application/json

HTTP/1.1 200 OK
Content-Type: application/json
{
  "invoice_id": "7f9c2d10-4b88-41e0-9a12-08d4c1109e33",
  "organization_id": 1042,
  "customer_name": "Alpha Manufacturing Example Corp",
  "amount_cents": 4850000,
  "remit_ach_account_last4": "8841"
}

Binding organization_id directly in the repository query—and returning 404 Not Found rather than 403 Forbidden so callers cannot probe foreign UUIDs—closes the lookup:

  invoice = db.session.execute(
-     select(Invoice).where(Invoice.id == invoice_uuid)
+     select(Invoice).where(
+         Invoice.id == invoice_uuid,
+         Invoice.organization_id == current_user.organization_id,
+     )
  ).scalar_one_or_none()

Sample Finding BD-2026-02: Unverified Payment Webhook Signature Allows Plan Activation

  • Severity: High (CVSS v3.1 8.1) · Class: CWE-345 (Insufficient Verification of Data Authenticity) · Endpoint: POST /webhooks/billing
  • Retest status: Remediated and verified on October 27, 2026 against commit 91c4e80b

Webhook endpoints have to stay reachable from the public internet, which makes cryptographic verification the entire perimeter. Here, the receiver parsed incoming JSON bodies and flipped enterprise subscription flags whenever type == "checkout.session.completed" without checking the Stripe-Signature HMAC header against raw request bytes. Enforcing raw-body signature verification with a 300-second replay tolerance and a unique event-ID idempotency table closed the issue.

Remediation Retest Table and Customer Letter of Attestation

Handing a prospective enterprise customer a static pentest PDF that still lists open Critical or High vulnerabilities tends to stall security reviews right away. Standards expect follow-through too: PCI DSS v4.0.1 Requirement 11.4.3 explicitly requires repeating tests to confirm that exploitable findings were corrected.

Once your engineers ship patches, the assessor replays every proof-of-concept request against the updated commit, appends a dated Retest Verification Table to the full technical report, and issues a standalone two-page Letter of Attestation. That letter summarizes scope, methodology, testing dates, and verified closure so you can satisfy vendor-risk questionnaires without handing out raw HTTP payloads or internal route names. If you need an engineer-led web application, API, or multi-tenant SaaS test with remediation retesting built in, see our penetration testing services.