1. Purpose and scope
This policy sets out how Carefully Coded Ltd uses independent penetration testing to check that the systems we build and run are secure. A penetration test is an authorised, simulated attack carried out by a qualified tester to find weaknesses before an attacker does.
It applies to:
- internet-facing systems we host or manage for clients, including web platforms, application programming interfaces (APIs) and mobile app back ends;
- software and devices we install on client sites, such as displays and sensor stations, where the contract or a risk assessment requires it;
- our own internet-facing company systems.
It forms part of our Information Security Policy. Where a client contract sets more frequent testing or a wider scope, the contract takes precedence.
2. When we test
We will commission an independent penetration test:
- at least every 12 months for each system in scope;
- after any major change, before the change goes live where possible and no later than 30 days after go-live;
- before a new client system first goes live, where the contract requires it.
A major change includes:
- a new system or a new public-facing component;
- a move to a new hosting provider, region or network design;
- significant changes to authentication, authorisation or session handling;
- new integrations that exchange personal or sensitive data;
- new device types or firmware that connect to client networks.
The director decides whether a change is major and records the decision in the change record (see our Vulnerability Management, Patching and Secure Development Policy). If in doubt, we treat the change as major.
3. Scope of each test
Before each test we agree and record a written scope with the tester and, where relevant, the client. The scope covers:
- the systems, URLs, IP addresses, apps or devices to be tested;
- the type of test, such as external infrastructure, web application, API, mobile app or device testing;
- testing windows, rules of engagement and emergency contacts;
- anything excluded, with a reason.
We test against non-production environments with representative data where possible. If production must be tested, we agree the timing with the client and make sure backups are current. Live personal data is not copied for testing except as allowed by our Test Data Policy.
Testing on client premises, such as airports, is only carried out with the client's written authorisation and in line with their site security rules.
4. Tester independence and qualifications
Tests must be carried out by a tester who is independent of the people who built and run the system. In practice this means an external specialist firm.
We will only use a tester who meets at least one of these criteria:
- the firm is a CREST member company, or the individual testers hold a current CREST qualification;
- the firm is a CHECK company approved by the National Cyber Security Centre (NCSC);
- the testers hold an equivalent, recognised industry qualification, agreed in advance with the client where the contract requires it.
Before engaging a tester we also check that they will sign a confidentiality agreement, hold suitable insurance and handle test data and reports securely. We choose a provider for each test against the criteria above. Suppliers who access client data are listed on our Sub-processors page where relevant.
5. Remediation
Each finding is logged in our issue tracker and fixed within the timescales in our Vulnerability Management, Patching and Secure Development Policy:
| Severity | Fix or mitigate within |
|---|---|
| Critical | 72 hours |
| High | 14 days |
| Medium | 30 days |
| Low | 90 days |
The clock starts when we receive the finding, which may be before the final report if the tester flags a serious issue during testing. If a finding cannot be fixed in time, we apply a mitigation and the director records a risk acceptance with a review date. If a test reveals signs of a real compromise, we start our Incident and Data Breach Response Procedure straight away.
6. Retesting
We ask the tester, or another tester meeting section 4, to retest:
- every Critical and High finding, within 30 days of the fix being deployed;
- Medium and Low findings at the next scheduled test, or sooner where the client requests it.
A finding is only closed when the retest confirms it is fixed, or when the director records a risk acceptance.
7. Reports and records
For each test we keep:
- the agreed scope and authorisation;
- the full test report;
- the remediation log, showing each finding, its severity, the action taken and the date closed;
- the retest report.
Reports are stored securely and access is limited to those who need it, because they describe weaknesses in our systems. We keep these records for at least the length of the client contract plus any period set out in our Data Retention Policy.
8. Sharing results with clients
Where a test covers a client's system, we will:
- tell the client of any Critical or High finding within 2 working days of receiving it, with our planned fix;
- provide the full report, or a summary if the client prefers, within 10 working days of a written request;
- provide the remediation log and retest results on request.
Requests should be sent to info@carefully.codes. Test results are reviewed at our annual management review, as set out in our Information Security Policy.