Our Team
COFFSec is a Hong Kong-based offensive security consultancy specialising in control validation, adversary emulation and technical investigation. Our consultants translate architecture-specific failure mechanisms into substantiated findings, actionable remediation and decision-relevant assessments of residual exposure.
Professional practice and delivery governance
01
Objective-led investigation
Engagement design begins with the decision requiring independent evidence. Release assurance examines transaction and tenancy boundaries; assumed-compromise assessment evaluates containment; incident investigation reconstructs historical activity. Each objective requires a distinct access model, validation method and evidential threshold.
Consultants establish the acting principal, expected control obligation, observed transition and consequential state change. Differential tests and corroborating evidence distinguish exploitable failure from environmental variation or an incorrect premise. The final explanation identifies the enforcement mechanism requiring correction.
02
Multidisciplinary accountability
Material attack paths commonly span application logic, workload identity and infrastructure administration. Specialist contributions retain their technical provenance, while the engagement lead remains accountable for scope coordination and the integrated conclusion. Delivery and review responsibilities are identified in the engagement plan.
- Engagement leadership governs scope, operational coordination and decision-support reporting.
- Application specialists validate authorisation semantics, input interpretation and transaction invariants.
- Infrastructure specialists reconstruct service exposure, delegated authority and compromise propagation.
- Code reviewers investigate implementation-level causes, framework semantics and deployment assumptions.
- Research contributors verify procedural prerequisites and safety constraints before client deployment.
03
Technical quality assurance
Technical review challenges causal reasoning, reachability, prerequisite validity and the proportionality of claimed impact. Reviewers assess whether remediation addresses the underlying defect class and whether another practitioner can reconstruct the conclusion from the evidence record.
Implementation defects, analytically feasible privilege paths and historical evidence support different claims. Reports preserve these classifications, including untested production reachability, laboratory-only demonstrations and incomplete incident chronology. Evidential confidence remains proportionate to the method and source quality.
04
Research integrity and reproducibility
Research specifies a falsifiable hypothesis, version-controlled baseline and controlled variables. Interpretation considers instrumentation artefacts, missing telemetry and competing explanations. Negative results retain analytical value when experimental conditions and limitations are documented accurately.
Publication is subject to confidentiality, intellectual-property rights and responsible-disclosure requirements. Client-derived material requires the governing agreement’s permission and any additional written approval before public research, training or commercial reuse.
05
Assessing engagement suitability
Selection should consider the proposed practitioners, comparable investigative experience and the quality of their methodological explanation. Credentials provide supporting context; relevant work samples demonstrate judgement under the technical and operational conditions of the intended engagement.
Proposed coverage, access requirements and residual uncertainty are explained before appointment. Alternative methods are considered where they would produce stronger evidence for the client’s decision, with specialist dependencies disclosed during scoping.
Delivery accountability and information-handling obligations are governed by the signed engagement agreement.
