Core principles

Inspect → decide → draw in one batch → look → fix → validate.

Skipping "look" is the single most common failure. You cannot tell from a tool result whether a sprite reads; you have to see it. look op preview shows you what a human sees; op ascii shows you exactly which pixel is where.

At 16×16 a character gets roughly 3 pixels of head, 5 of torso and 4 of legs. There is no room for a nose. Decide what the sprite must communicate at a glance — a class, a threat, a direction — and spend the pixels there.

Common sizes and what fits:

Size	Fits
8×8	An icon or an item. One idea.
16×16	A readable character with 2–3 colours per material.
32×32	Facial suggestion, distinct armour pieces, believable weapon.
48×48+	Real detail; now the risk is noise, not scarcity.

Pick and write down, before drawing: canvas size, palette, light direction, outline style, and whether the sprite is side-on, top-down or 3⁄4. Every later decision either follows from those or is a mistake.

In rough order of how quickly a person spots it:

1. Too many colours. Ninety near-identical browns instead of a four-step ramp. Fix: a palette, and paletteLock left on.
2. Flat luminance shading. Shadow = the same hue, darker. Real shading shifts hue (see rules://02-shading-and-light).
3. Anti-aliased or semi-transparent pixels. Usually from a resize or a soft brush. At 1× they read as blur, not smoothness.
4. Noise. Isolated pixels with no neighbour, "detail" that is just dirt.
5. Mushy silhouette. The shape does not read as anything when filled black.
6. Mechanical timing. Every animation frame the same duration.

validate checks 1, 3, 4 and 6 mechanically. The rest you have to look at.

Do not resize their canvas, re-index their colour mode, flatten their layers or overwrite their file unless they asked. When a task seems to require one of those, say so and ask.