Help Centre
This guidance addresses assessment commissioning, access preparation, operational coordination and interpretation of assurance evidence. Initial consultation establishes the decision requirement and material uncertainty; detailed specifications and sensitive implementation records are requested only where necessary for scoping.
Practical guidance for an engagement
01
Preparing the first conversation
An initial brief should identify the business capability, relevant stakeholders and commissioning rationale. Release approval, supplier assurance and suspected intrusion require different evidential approaches. Consultation translates the stated uncertainty into assessable objectives, proportionate coverage and realistic access prerequisites.
- Identify the systems, important technologies and approximate scale.
- State whether testing would involve production, staging or both.
- Describe the roles or tenants whose separation needs to be examined.
- Mention deadlines, prohibited activity and information handling requirements.
02
How an engagement progresses
The proposal establishes objectives, coverage and deliverables. Rules of engagement specify executable operating boundaries, including permitted procedures, timing, escalation and stop authority. Material amendments require prior approval; asset discovery does not independently extend the authorised scope.
- Scoping establishes the objective, asset authority and available access.
- Preparation confirms account readiness, provider restrictions and escalation contacts.
- Testing records evidence within approved limits, with urgent exposure escalated promptly.
- Review resolves factual questions before the final report is issued.
- Handover explains the result, remediation priorities and any agreed retest.
03
Providing representative access safely
Test accounts should reflect the real permission boundaries, including different roles or tenants where relevant. Synthetic records with known ownership allow an access-control failure to be demonstrated without browsing unrelated customer content. Account access does not grant authority to test every service it can reach.
Document material differences between staging and production. An alternate identity provider, feature flag or integration can change the result. Before active work, nominate a contact who can pause testing and confirm how unexpected instability will be handled.
04
Reading a finding in context
Interpret findings against their starting privileges, tested version and exploitation prerequisites. Administrator-assisted and unauthenticated outcomes represent different exposure scenarios. Technical scoring characterises selected vulnerability properties; business treatment priority additionally reflects asset criticality, reachable attack paths and compensating controls.
The coverage record is equally important. An inaccessible function was not necessarily protected; it may simply have remained untested. A focused retest checks the original condition with relevant bypass cases. Newly added functionality or a different architecture may require separate scope.
05
Requesting a correction or clarification
Provide the finding reference, tested version and precise point in dispute. A changed deployment, a factual error and a disagreement about business priority require different responses. Relevant configuration evidence helps establish which applies. Request a protected channel before sharing confidential material.
We compare new information with the original evidence. Confirmed errors are corrected with version context. A supported finding is not removed solely because it is inconvenient, and any unresolved disagreement is documented with its technical basis.
06
Reporting a concern about our systems
Begin with a minimal description of the affected component and the observed impact. Do not access another person’s records, establish persistence or continue testing beyond what is necessary to describe the concern. Ask for a protected route before sending sensitive evidence. This guidance is not authorisation to test unrelated assets.
Reports are normally acknowledged within two business days. Investigation time depends on the issue and available evidence. An acknowledgement confirms receipt; it does not mean the vulnerability has been verified.
07
Response expectations
- New scoping enquiries are normally answered within one business day.
- Active-engagement communications follow the agreed channel and named delivery lead.
- Incident availability must be confirmed or activated under an applicable retainer.
- Concerns about our own systems are normally acknowledged within two business days.
Business days are Monday to Friday in Hong Kong time. Support outside those hours follows the applicable agreement.
08
Choosing the next step
Compare services by the evidence needed for the decision. Runtime testing and source review can complement each other before a release. A suspected intrusion may require preservation before any configuration change. Procurement requirements should be raised early, because contractual review or provider permission can affect the start date.
Specific commitments are governed by the signed agreement, not by this general guide.
