Methodology

Beyond Automated Scanning. Security Testing That Delivers Actionable Results.

Our methodology is built around one principle: automated and AI-assisted tools help us find candidate issues faster, but only a human tester decides what is real. Here is exactly how an engagement runs, from the first conversation to verified fixes.

Six phases

How Every Engagement Runs

  1. Phase 1

    Scoping

    Agree exactly what will be tested, how and when — in writing, before any testing begins.

    • Identify authorized targets: domains, URLs, APIs, repositories and environments
    • Understand the application architecture, technology stack and key data flows
    • Define testing boundaries, including what is explicitly out of scope
    • Identify authentication roles and arrange test accounts for each
    • Agree testing windows, exclusions, contacts and emergency stop procedures
  2. Phase 2

    Discovery

    Build a detailed picture of the attack surface before testing in depth.

    • Application mapping by navigating every in-scope feature as each role
    • Endpoint discovery from documentation, client code and observed traffic
    • Technology identification: frameworks, libraries, hosting and third-party services
    • Attack surface identification: inputs, uploads, integrations and trust boundaries
    • AI-assisted vulnerability discovery to highlight candidate weaknesses quickly
  3. Phase 3

    Testing

    Hands-on testing of the areas most likely to hide serious vulnerabilities.

    • Authentication testing: login, MFA, password reset, sessions and tokens
    • Authorization testing: horizontal and vertical access control across roles
    • API security testing: object- and function-level authorization, data exposure
    • Input validation testing: injection, XSS, SSRF, file handling and deserialisation
    • Business logic testing: workflows, limits, race conditions and abuse cases
    • Source code analysis where code access is part of the agreed scope
  4. Phase 4

    Validation

    Nothing is reported as a confirmed vulnerability until a person has proven it.

    • Reproduce each potential vulnerability manually in a controlled way
    • Eliminate false positives, duplicates and theoretical-only issues
    • Assess real security impact in the context of your application and data
    • Collect appropriate evidence — requests, responses and screenshots — while minimising data exposure
  5. Phase 5

    Reporting

    A clear, structured report for both decision makers and developers.

    • Executive summary of overall risk, key themes and priorities
    • Technical findings with affected components and evidence
    • Severity classification using CVSS with business-context notes
    • Step-by-step reproduction instructions
    • Practical remediation guidance and defence-in-depth recommendations
  6. Phase 6

    Retesting

    Confirm that fixes work, and document the updated position.

    • Verify developer fixes against the original reproduction steps
    • Try reasonable bypasses of each fix
    • Record remaining or partially fixed issues
    • Deliver updated findings showing current status

Our use of AI

AI Is an Assessment Aid, Not the Assessor

We use AI-assisted security tooling to accelerate the parts of testing that are repetitive or large in volume: mapping applications, enumerating endpoints, generating test cases, triaging scanner output and highlighting code paths that deserve a closer look.

AI output can be wrong. It can report issues that do not exist and miss issues that do. That is why every potential vulnerability identified by automated or AI-assisted means is manually investigated and validated before it appears in your report as a confirmed finding. Where something looks suspicious but cannot be confirmed, we say so clearly rather than presenting it as fact.

The result is the efficiency of modern tooling with the accuracy and judgement of a human tester — which is how we keep assessments affordable without cutting corners.

AI-assisted tooling helps with

  • Speed and breadth of discovery
  • Consistent coverage of repetitive checks
  • Spotting patterns across large codebases and APIs
  • Prioritising where a human should look first

A human tester is responsible for

  • Understanding business context and intent
  • Confirming exploitability and impact
  • Rejecting false positives
  • Every finding, rating and recommendation in the report

Rules of engagement

Safe, Authorised and Confidential

Written authorisation first

We only test systems you own or are authorised to have tested, after scope and authorisation are agreed in writing. An enquiry is never treated as permission to test.

Controlled, non-destructive testing

Tests are designed to prove a vulnerability with minimal impact. Denial-of-service and destructive testing are excluded unless explicitly agreed with safeguards.

Confidential handling

Findings, evidence and any access you provide are treated as confidential, shared only with your nominated contacts and retained only as long as agreed.

We use recognised community references — including the OWASP Web Security Testing Guide, OWASP Top 10, OWASP API Security Top 10, OWASP ASVS and CVSS — to structure testing and classify findings. These are methodological references, not certifications.

Get started

See the Methodology in Action

Tell us about your application and we'll show you how we would approach it — scope, timeline and deliverables.

  • A scoping conversation to understand your application
  • A written proposal before any work starts
  • Testing only with written authorisation

Or write to aponlineventures@gmail.com

Quick enquiry

Please don't include passwords, API keys, source code or confidential data. We'll arrange secure handover of anything sensitive after scoping.

Prefer email? aponlineventures@gmail.com

Submitting this form does not authorise any security testing. Testing only begins after scope and written authorisation are agreed. We never scan a URL just because it was submitted here.