ComicBookGenerationTask
Generate structured comic book scripts and high-fidelity visual strips with character consistency and multi-page orchestration.
Category: Writing
Multi-Modal
DALL-E / Midjourney Integration
⚙️ ExecutionConfig.json
{
"subject": "Cyberpunk Detective in Neo-Tokyo",
"target_pages": 5,
"art_style": "manga",
"style_details": "High contrast, neon accents",
"generate_images": true
}
→
👁️ Session UI Output
PAGE 1: THE RAIN FALLS
[NEO-TOKYO STREETSCAPE]
Kaito: "The data is here. Somewhere in the neon haze."
[CLOSE UP: GLOWING DATA CHIP]
Generated: comic_book.json, 15 images
Live Results Showcase
Explore actual artifacts generated by this task, including character sheets and page strips.
Execution Configuration
| Field | Type | Description |
|---|---|---|
subject * |
String | The subject or scenario to develop into a comic book. |
target_pages |
Int | Target number of pages (Default: 5). |
art_style |
String | Art style (e.g., 'manga', 'noir', 'cartoon'). |
style_details |
String | Additional visual guidelines or lighting details. |
generate_images |
Boolean | Whether to generate images for each row (Default: true). |
Token Usage: High (Multiple LLM calls for script + Image generation tokens).
Task Lifecycle
- Script Generation: Uses a smart model to create a
ComicScriptobject containing characters, pages, rows, and frames. - Character Profiling: Extracts visual traits for each character to ensure consistency.
- Reference Generation: Generates character sheets to be used as visual context for subsequent panels.
- Strip Generation: Iterates through pages and rows, generating composite images for each "strip" using character references and scene descriptions.
- Finalization: Packages the script and image links into a
comic_book.jsonworkspace file.
Embedded Execution (Serverless)
Use the UnifiedHarness to run this task programmatically in CI/CD or CLI tools.
import com.simiacryptus.cognotik.plan.tools.writing.ComicBookGenerationTask.Companion.ComicBookGeneration
import com.simiacryptus.cognotik.plan.tools.writing.ComicBookGenerationTask.ComicBookGenerationTaskExecutionConfigData
// 1. Define the Job
val executionConfig = ComicBookGenerationTaskExecutionConfigData(
subject = "A space western about a lonely robot",
art_style = "western superhero",
target_pages = 3,
generate_images = true
)
// 2. Run via Harness
harness.runTask(
taskType = ComicBookGeneration,
typeConfig = TaskTypeConfig(), // Default static settings
executionConfig = executionConfig,
workspace = File("./my-comic-project"),
autoFix = true // Skip manual approval for image generation
)
CLI / JSON Invocation
# Example of how the task is identified in the orchestration engine
cognotik-cli --task ComicBookGeneration \
--input '{"subject": "Cyberpunk Noir", "art_style": "manga"}'
LLM Prompt Segment
ComicBookGeneration - Generate comic book scripts and visuals
- Create a comic book script with page/row/frame structure
- Specify subject, target pages, and art style
- Generates character profiles and visual descriptions
- Can generate images for each row (strip)
- Create a comic book script with page/row/frame structure
- Specify subject, target pages, and art style
- Generates character profiles and visual descriptions
- Can generate images for each row (strip)