A.S.M. is the harm scoring engine at the center of C.A.S. It looks at a platform's design choices, how much content it pushes, how aggressively its algorithm amplifies that content, how emotionally charged the content is, how fast engagement loops move, and how strong social feedback pressure is, and turns those design choices into a single risk score. Unlike other scores in C.A.S. that go up when things are healthy, A.S.M. runs in reverse, meaning the higher the score, the more harmful the platform is to attention.

A.S.M. traces a direct chain, it looks at a platform's design pressures, converts that into an overall pressure reading, then maps that pressure onto a harm level ranging from low to critical. Two things can shift the result, sustained volatility in attention load pushes harm higher, and strong recovery mechanisms can bring harm down slightly, but only when a platform is already in the safer range. A platform already flagged as severe or critical cannot buy its way down with recovery features alone.

As pressure rises, A.S.M. also tracks how people respond to it, a predictable sequence running from selective attention, to muting, to avoidance, to disengagement, to full withdrawal. These are the human reactions to an unsafe attention environment, not settings a platform controls.

The platform operates through orchestration logic that triggers multiple compliance engines in sequence to process assessments and finalize report data. The orchestration layer runs functions such as evidence intelligence, multi state compliance, and legal citation analysis, then aggregates all results into a master report for consistent display across the platform.

At the center of this workflow is the Master Report, which consolidates results from scoring, compliance, and governance subsystems into a unified entity. The evidence engine then dynamically pulls from raw platform crawls, user uploaded evidence, and statutory profiles to determine whether system mechanisms meet legal and policy requirements.