Following assumes one has a general idea how severe weather gets going, background theory. Also assumes one has their favorite data/model websites already. Below is a very basic few keys to look for operationally.
A system: Look for surface low pressure with a trailing trough on surface charts/models. A short-wave aloft is required. One can look at 500 mb PVA, 200/300 mb winds, 700 mb short-waves, and others.
Instability: One can look at model CAPE but I'm not a fan in the extended period. I look for temps and dewpoint forecasts. 75/60 is decent right now. Like 80/65 in April. Prefer 85/70 in May none are exactly. Seen some 80/68 go nuts in May.
Wind shear: Look for speed and directional shear forecast. Very basic I want a low level jet at 850 mb 30-40 knots from the south. Look for southwest or west winds at 700/500/200/300 mb. Best if each level up is more westerly, turning with height.
Models will try to get at helicity/SRH a few days out, but I really prefer looking at each level. Even closer to event time models have instability/wind shear combos like EHI - energy helicity index. I still drill down the charts myself.
Same day those indices (EHI/SRH) add value over just eyeballing the charts. Models have equations that pick up on some vorticity same day. High-res models are designed for severe weather, so I start trusting the radar simulations 12-36 hours out.
Still, I forecast myself while using models. Divergence reduces confidence. Agreement boosts confidence.
Many other specifics require consideration, but above is a start. Each level has preferred levels of temperature, moisture, divergence, convergence, height falls, vertical motion, and more. However in classic set-ups these tend to work out.
I would focus on the Big 3: a system, instability, and wind shear.