# AIE019-LAB1 — Side-node Power Orchestration

Implement `evaluate_power_intent(case)` in `starter.py`. Keep desired state, observed state, and runtime readiness separate. Your result must contain these eight fields: `evidence_fresh`, `control_available`, `timing_ok`, `shutdown_eligible`, `wake_eligible`, `transition_state`, `final_decision`, and `scheduler_admission_state`.

Use only the supplied synthetic fixtures. A control acknowledgement must never be treated as observed state or readiness proof. `ForceOff` is not a normal optimization path. Run `python run_checks.py` to verify fixture structure; the hidden evaluation checks deterministic decisions.