Skip to main content

SENTRIX / Service overview

Evidence that an attack works,
and a clear standard for the fix

From access control in web apps and APIs to the data flows of AI features.
We provide real reproduction evidence and a direction for the fix.

3-page PDF · for internal review and sharing

When to review this

Before a launch · after changes to accounts or permissions · when adding AI features · when managing external assets

SENTRIXSECURITY REVIEW
Deliverable preview · synthetic example

Testing access control
between customer accounts

Test account AReport APIOrganization B's data
01 / Evidence
Accounts, requests, responses and reproduction conditions
02 / Fix direction
Server-side checks of organization and resource ownership
03 / Retest
Confirm access is blocked after the fix

An example to explain the report layout, not a real finding from a customer system

01 / SCOPE

Scope set per target,
based on what you need to validate

The scope follows your target systems and business flows.

01

Web and API penetration testing

Before a launch, after changes to the permission model, or when checking important business flows

  • Data access between users and organizations
  • Admin functions, authentication and session handling
  • Whether business rules such as payments and settlement can be bypassed
How we assess
02

AI service security assessment

When you introduce or expand LLM, retrieval-augmented or AI agent features

  • Prompt input and injection through external documents
  • Access control and sensitive data exposure in retrieved material
  • Tool execution permissions and how far data is passed on
Assessment scope
03

External exposure and risk management

When you manage your organization's internet-facing assets together with follow-up fixes

  • Discovering external assets and confirming ownership evidence
  • Tracking priorities, owners and fix status
  • Recheck history and reporting material
About ASM

AI-driven in-depth vulnerability assessment combines automated analysis with expert validation. Standard web apps and APIs that do not use LLMs can also be assessed.

02 / DELIVERABLES

Material for internal reporting,
and evidence for your developers' fixes

Confirmed results are kept separate from items that need further validation.
A synthetic example to explain the report layout

Example F-01 / Access control

When one organization can open another organization's reports

Synthetic example
Reproduction condition
Log in with a test account of organization A, then request a report ID that belongs to organization B.
What we check
Whether the server verifies ownership, with the request, response, role and impact recorded. Sensitive data in raw evidence is masked.
Fix direction
Verify the user's organization and the resource owner together on the server, and apply the same policy to list, detail and download paths.
Before and after the fix
Before the fix, another organization's test data could be accessed. After the fix, we confirm that access is blocked while users can still view their own data. A synthetic example to explain the report layout.
Open items
APIs or account roles we did not test are marked as unverified. A confirmed fix for this target is not stretched into a claim that the whole service is safe.
Security and decision makers

Risk summary and priorities

The business affected, the scope confirmed, the remaining risk and the order of fixes.

Developers and operators

Reproduction evidence and fix direction

Reproduction conditions and steps, evidence, fix recommendations and the results of the agreed retest.

Compare deliverables by service A blank sheet for comparing several quotes on the same scope (printable)

03 / DELIVERY

From the agreement before kickoff
to the check after the fix.

Operational impact and data-handling terms are decided together with the scope.

  1. 01

    Target and scope consultation

    We confirm the service, account roles, business flows to test and your preferred schedule.

  2. 02

    NDA and written approval

    We agree on targets and exclusions, allowed hours, access methods and who to call to stop testing.

  3. 03

    Assessment and results review

    We test the agreed scenarios and explain the evidence, impact and fix direction.

  4. 04

    Fixes and retest

    We confirm the fix results and remaining risk within the period and number of retests set in the contract.

Data handling

Who has access, retention and deletion, whether external AI or APIs process data, and how far data is passed on are confirmed before the contract. Your first inquiry only needs the target and purpose, with no confidential information.

Cost and schedule

We send a proposal after checking the number of systems, account roles, APIs and business flows, the report format and the retest scope. ASM is quoted after consultation.

Who assesses, who fixes

SENTRIX assesses, reviews the evidence and explains the fix direction. Your developers and operators review and apply the fixes. Fixing on your behalf and device changes are agreed separately, including who does them and their scope.

How results are reviewed

Confirmed results, candidates that need further checks and excluded scope are kept separate. Retests compare before and after the fix and the remaining open items, within the contracted scope.

Roles and track record

04 / SENTRIX

Validation from the attacker's point of view,
and follow-up fixes in the field

SENTRIX provides security assessments and external asset management.
We also run our own security check service, CodeScan.

Team experience10 years in security
Team awardTop Excellence Award, 2025 AI Red Team Challenge hosted by Korea's Ministry of Food and Drug Safety
Vulnerability reportsReported to Korea's national vulnerability database (KVE)

Experience, awards and reports of our team members. Mentioning an organization does not imply a partnership, sponsorship or customer relationship. More about our track record

LET’S TALK

Talk to us about the service
and your preferred schedule

We reply within one business day

Sentrix · CEO Kim Soha · Business Registration No. 673-54-01057

Download the company profile (PDF, Korean) ↓