Performance and Diagnostics

This policy chooses which AI decisions can run based on system conditions. It allows AI behavior to slow down predictably. It does not use unpredictable timers. It keeps safety rules intact.

This service decides when approved AI tasks can run. It uses set times, triggers, and updates instead of random checks. It keeps budget limits and data safety rules. The system controls when AI work starts based on permissions and load.

This service validates diagnostic events. It controls their cost and groups repeated events. It prepares projections for reports or analysis. It does not write data to disk or show it on screens during live operation.

This service records timing and pressure facts for Brain to enforce runtime budgets. It uses user-mode timing and bounded event records. It does not use privileged CPU counters, disk logging, or string-heavy diagnostics. The service only supports minimal telemetry for scheduler and budget governance. Kestrel presentation is not part of this service. Broader diagnostic aggregation or retention is not handled here. It cannot slow down execution or destabilize the hot path. It only records facts that Brain needs.

This policy says how Brain handles non-critical tasks when system pressure threatens live operations. It makes sure safety and integrity work always runs. Deferred or degraded results must be clear to callers and diagnostics. The system does not hide lower fidelity outcomes. Brain only reduces or delays work that is not required for safety. This policy protects the core functions of the system.

The Runtime Budget Governor is a Brain service. It manages runtime budgets by watching live timing and pressure signals. It sets budget states for scheduler, plugin runtime, and other hot-path services. It protects the host-facing runtime. It decides when work must continue, yield, defer, or degrade. The Brain owns budget state. It does not use universal numeric limits. It only uses observed runtime cadence and approved policy. This service does not make decisions for other services. It only publishes budget facts. It does not control plugin operations directly. It does not use jargon without explanation.

ADK divides runtime work into protected and deferrable budget families so one expensive feature cannot consume the whole game tick. When spare processor capacity and safe parallel work are available, Brain may add bounded worker assistance before queues overflow.

For example, a burst of NPC planning may use an extra worker on an idle core, while accepted mutation receipts and host-critical work retain their protected authoritative-thread budget.