refactor: simplify CLAUDE.md, move detailed docs to docs/

CLAUDE.md reduced from 621 to 76 lines (↓88%). Detailed guidelines
moved to on-demand reference files under docs/.

docs/image-generation.md

@@ -0,0 +1,53 @@
+# Image Generation Guidelines
+
+Skills that require image generation MUST delegate to available image generation skills.
+
+## Skill Selection
+
+**Default**: `skills/baoyu-image-gen/SKILL.md` (unless user specifies otherwise).
+
+1. Read skill's SKILL.md for parameters and capabilities
+2. If user requests different skill, check `skills/` for alternatives
+3. Only ask user when multiple viable options exist
+
+## Generation Flow Template
+
+```markdown
+### Step N: Generate Images
+
+**Skill Selection**:
+1. Check available skills (`baoyu-image-gen` default, or `baoyu-danger-gemini-web`)
+2. Read selected skill's SKILL.md for parameters
+3. If multiple skills available, ask user to choose
+
+**Generation Flow**:
+1. Call skill with prompt, output path, and skill-specific parameters
+2. Generate sequentially by default (batch parallel only when user has multiple prompts)
+3. Output progress: "Generated X/N"
+4. On failure, auto-retry once before reporting error
+```
+
+**Batch Parallel** (`baoyu-image-gen` only): concurrent workers with per-provider throttling via `batch.max_workers` in EXTEND.md.
+
+## Output Path Convention
+
+**Output Directory**: `<skill-suffix>/<topic-slug>/`
+- `<skill-suffix>`: e.g., `xhs-images`, `cover-image`, `slide-deck`, `comic`
+- `<topic-slug>`: 2-4 words, kebab-case from content topic
+- Conflict: append timestamp `<topic-slug>-YYYYMMDD-HHMMSS`
+
+**Source Files**: Copy to output dir as `source-{slug}.{ext}`
+
+## Image Naming Convention
+
+**Format**: `NN-{type}-[slug].png`
+- `NN`: Two-digit sequence (01, 02, ...)
+- `{type}`: cover, content, page, slide, illustration, etc.
+- `[slug]`: 2-5 word kebab-case descriptor, unique within directory
+
+Examples:
+```
+01-cover-ai-future.png
+02-content-key-benefits.png
+03-slide-architecture-overview.png
+```