0.11.0 — vocabulary 23: stickman, the default avatar

Stickman, the default avatar: a new core primitive stickman (spec 030), an articulated stick
figure that acts, gestures at the rest of the frame, speaks, moves with host data, and draws a
whole crowd from one layer. No existing reference frame moved.

• 15 actions: idle, walk, run, jump, punch, kick, hit, fall (tips over once and stays down),
  ko, wave and dance, plus point, present, reach and look aimed at
  targetX/targetY; travel and facing. Every pose is a closed-form function of time.
• Interaction through params and frame data: aim at a target such as a line-chart's newest
  point, speech (the frame's label in a bubble), from: "coords" (stand where the host
  says), and drive level or beat; every number is bindable.
• Crowds in one layer: count up to 12, arrangement row, cluster or split, each figure's
  phase and size from the seed, one highlight and one speaker, coords[index + i] per
  figure.
• Six styles (classic, bold, outline, chibi, pixel, joints) and colour (fg, accent, fg2).
  The pixel style, and every style under pixelMode, is exact in the palette and the
  profiles.
• In the core vocabulary and the data-viz, minimal and arcade kits. Showcase:
  stickman-presenter, stickman-fight and stickman-dance; a gallery demo.
• AI prompting: three stickman use cases (data narrator, scenes from a prompt, interactive);
  the vocabulary example now matches the select / selectDetailed API.
• Dev dependency: @types/node 26.6.6.