Skill v1.0.0
currentLLM-judged scan90/100version: "1.0.0" name: blog-image-prompts description: "Generate AI image prompts and placement points for a blog post."
Blog Post Image Prompt Generator
Workflow Checklist
Copy and track progress:
Image Prompt Progress:- [ ] Step 1: Analyze blog post (type, tone, themes)- [ ] Step 2: Identify placement points- [ ] Step 3: Determine visual style- [ ] Step 4: Generate prompts- [ ] Step 5: Validate prompts- [ ] Step 6: Write output file
Process
Step 1: Analyze the Blog Post
Identify:
- Content type: Technical, personal essay, business, opinion, how-to
- Key themes: Concepts, metaphors, or ideas that could be visualized
- Structure: Sections and natural break points
Step 2: Identify Placement Points
| Location | When to Use | |
|---|---|---|
| Hero | Almost always | |
| Section breaks | Long posts with distinct sections | |
| Concept illustrations | Abstract ideas needing visualization | |
| Emotional beats | Key moments in narratives |
Quantity (flexible guidelines):
- Short posts: 1-2 images
- Medium posts: 2-4 images
- Long posts: 3-6 images
Not every section needs an image.
Step 3: Determine Visual Style
| Content Type | Style Direction | |
|---|---|---|
| Technical | Diagrams, isometric, flat design | |
| Personal essays | Editorial photography, painterly, atmospheric | |
| Business | Professional, polished | |
| Opinion | Conceptual, metaphorical, abstract |
Step 4: Generate Prompts
Each prompt must include these six elements:
- Subject - Main focus
- Style - Illustration, photograph, diagram, etc.
- Mood - Lighting, color palette, emotional tone
- Composition - Framing, perspective, negative space
- Details - Specific elements from the blog content
- Alt text - Descriptive alt text for the generated image (for accessibility and SEO)
For concrete examples, see references/prompt-examples.md.
Step 5: Validate Prompts
Before presenting, verify each prompt:
- [ ] References specific content/metaphors from the post
- [ ] Style is consistent across all prompts
- [ ] Aspect ratio specified (16:9 for headers, 1:1 or 4:3 for inline)
- [ ] Explicitly excludes common AI clichés (lightbulbs for ideas, handshakes for business, etc.)
- [ ] No text rendering requested (AI struggles with text)
- [ ] Alt text is descriptive and contextual (not just restating the prompt)
Revise any prompt that fails these checks.
Step 6: Write Output File
Write the prompts to a markdown file for collaborative editing.
File naming: Save alongside the blog post as [blog-post-name]-image-prompts.md
Example: If the blog post is draft-the-performance-treadmill.md, create draft-the-performance-treadmill-image-prompts.md in the same directory.
File format:
# Image Prompts for "[Blog Post Title]"**Content analysis**: [1-2 sentence summary]**Recommended image count**: [N] images**Visual style direction**: [Overall style]---## Image 1: [Title]**Placement**: [Location in post]**Purpose**: [What it accomplishes]**Prompt**: [Full prompt text]**Alt text**: [Descriptive alt text for accessibility]**Style notes**: [Aspect ratio, what to avoid]---## Image 2: [Title]...
After writing the file, confirm the file path to the user so they can open it for review and editing.
Flexibility Guide
- High freedom: Image count, exact placement, style interpretation
- Medium freedom: Output format structure (maintain organization, adapt sections)
- Low freedom: Six prompt elements (always include all six), validation checks