01 / Section
Focus on tasks instead of employee scores
A workflow can explain how a client renewal, report, or support resolution moved through systems. That evidence does not justify a conclusion about someone’s diligence, ability, or value.
Shadow is designed around task-level understanding: outcomes, steps, sources, and recurring processes. Employee scoring and sensitive personal inference sit outside that boundary.
02 / Section
Account for gaps in captured activity
Capture gaps, excluded windows, offline periods, and work completed outside the observed device all affect what the record can show. A responsible analysis describes the evidence it has without pretending the observed span is the whole working day.
Coverage should qualify a claim, not become a proxy for effort or productivity.
03 / Section
Scope access by verified identity
Cloud data belongs to an authenticated organization, user, and device scope. The service derives that boundary from verified sessions or enrolled-device credentials—not from a request field or a convenient query parameter.
Review access is also distinct from ordinary workspace membership. Keeping those checks explicit reduces the chance that sensitive context reaches the wrong audience.
04 / Section
Give people clear capture controls
People can pause capture and use supported exclusions for displays, window titles, and browser URL patterns. Reviewers can follow claims to evidence and decide whether an interpretation holds.
Shadow organizes evidence and shows possible opportunities. People make the final decisions.