Add Clawd photo compositor skill

[xixicc186]

What

• add the clawd-photo-compositor skill with 16 bundled Clawd pose references
• add pose-selection metadata and reference-guided compositing instructions
• document installation and usage in both Chinese and English

Why

This gives the repository a second Clawd workflow: adding the canonical flat 2D Clawd character to user photographs while matching object composition and interaction.

Behavior safeguards

• preserve Clawd's reference proportions, flat geometry, block hands, and canonical colors
• treat gray reference geometry as a pose proxy only, so it is not copied into the output
• preserve the source photograph outside the requested interaction

Validation

• quick_validate.py skills/clawd-photo-compositor — passed
• pose index contains 16 entries and every referenced asset exists