← Back to Skills Marketplace
nitishgargiitd

Comi Cog

by CellCog · GitHub ↗ · v1.0.11 · MIT-0
darwinlinuxwindows ✓ Security Clean
2353
Downloads
3
Stars
5
Active Installs
12
Versions
Install in OpenClaw
/install comi-cog
Description
AI comic and manga creation powered by CellCog. Comics, manga, webtoons, graphic novels, comic strips, sequential art. Character-consistent panels and visual...
README (SKILL.md)

Comi Cog - Comics & Manga Powered by CellCog

Create visual stories with AI - from manga pages to webtoons to comic strips with consistent characters.

How to Use

For your first CellCog task in a session, read the cellcog skill for the full SDK reference — file handling, chat modes, timeouts, and more.

OpenClaw (fire-and-forget):

result = client.create_chat(
    prompt="[your task prompt]",
    notify_session_key="agent:main:main",
    task_label="my-task",
    chat_mode="agent",
)

All agents except OpenClaw (blocks until done):

from cellcog import CellCogClient
client = CellCogClient(agent_provider="openclaw|cursor|claude-code|codex|...")
result = client.create_chat(
    prompt="[your task prompt]",
    task_label="my-task",
    chat_mode="agent",
)
print(result["message"])

Why Comi-Cog is Complex Work

Comics are one of the most demanding creative outputs:

  • Character Consistency: Same character must look identical across dozens of panels
  • Visual Storytelling: Every panel needs composition, flow, and meaning
  • Sequential Art: Panels must read naturally, guide the eye, create rhythm
  • Text Integration: Speech bubbles, sound effects, narration boxes
  • Style Coherence: Art style must stay consistent throughout

This is complex work. CellCog excels here because it maintains context across panels and pages, ensuring your characters look like themselves in every frame.


What Comics You Can Create

Manga Pages

Japanese-style sequential art:

  • Action Manga: "Create a fight scene manga page with dynamic movement"
  • Slice of Life: "Make a cozy manga page of friends at a café"
  • Shonen Style: "Create an intense rivalry moment between two characters"
  • Romance: "Make a confession scene in shoujo manga style"

Example prompt:

"Create a manga page (4 panels):

Scene: Hero confronts the villain for the first time

Panel 1: Wide shot - hero enters dark throne room Panel 2: Close-up - villain's smirk from shadow Panel 3: Dramatic - villain stands, revealing full design Panel 4: Reaction - hero's determined expression

Style: Dark fantasy shonen (like Berserk meets Demon Slayer) Include: Speed lines, dramatic shadows, Japanese SFX

Characters:

  • Hero: Young warrior, silver hair, scar across eye, armored
  • Villain: Elegant, long black hair, flowing robes, unsettling beauty"

Webtoon Episodes

Vertical scrolling format:

  • Vertical Strips: "Create a webtoon episode in vertical scroll format"
  • Cliffhangers: "Make a webtoon ending that hooks readers"
  • Romance Webtoon: "Create a sweet moment between the leads"
  • Action Webtoon: "Design a chase scene for vertical reading"

Example prompt:

"Create a webtoon episode (vertical format, 8-10 panels):

Story: Fantasy romance - a witch and a knight meet for the first time

Flow:

  • Knight lost in enchanted forest
  • Discovers cottage covered in flowers
  • Meets the witch (comedic first impression - she's not what he expected)
  • End on her mysterious smile

Style: Soft colors, romantic fantasy, clean line art Format: Vertical webtoon (panels flow downward)"

Comic Strips

Newspaper-style short form:

  • Daily Strips: "Create a 4-panel comic strip about office life"
  • Gag Comics: "Make a 3-panel joke about cats"
  • Webcomic Style: "Create a comic strip in the style of xkcd"
  • Sunday Comics: "Design a larger format weekend comic strip"

Example prompt:

"Create a 4-panel comic strip:

Setup: Programmer finally fixes a bug Punchline: Creates three new ones in the process

Style: Clean, simple, relatable (like Dilbert meets modern tech humor)

Include expressions that sell the emotional journey: Panel 1: Frustration Panel 2: Determination Panel 3: Triumph Panel 4: Dawning horror"

Graphic Novel Pages

Full-format sequential art:

  • Chapter Pages: "Create the opening page of a graphic novel chapter"
  • Splash Pages: "Design a dramatic full-page spread"
  • Dialogue Scenes: "Make a character conversation page with interesting staging"
  • Action Sequences: "Create a two-page action spread"

Character Consistency

The magic of comi-cog: your characters stay consistent.

When you describe a character, CellCog maintains their appearance across all panels:

Good character description:

"Character - Luna:

  • Age: Early 20s, petite build
  • Hair: Long silver hair with bangs, usually in a braid
  • Eyes: Large, purple, expressive
  • Outfit: Dark blue witch robes with star embroidery
  • Distinguishing: Small mole under left eye, always wears moon earring
  • Expression range: Usually serious but has a warm smile"

What this enables:

  • Same face structure across all panels
  • Consistent outfit details
  • Recognizable from any angle
  • Emotional range while staying "her"

Comic Styles

Style Characteristics Best For
Manga Expressive eyes, speed lines, screen tones Action, romance, drama
American Comics Bold lines, dynamic poses, vivid colors Superheroes, action
Webtoon Clean lines, soft colors, vertical flow Romance, slice of life
Indie/Alt Unique art styles, experimental Personal stories, art comics
Webcomic Simple, expressive, quick read Humor, daily updates
Graphic Novel Detailed, painterly, cinematic Literary, mature themes

Page Layouts

Request specific layouts:

Layout Panels Use Case
Grid 4-6 equal panels Steady pacing, dialogue
Asymmetric Mixed sizes Emphasis and flow
Splash Full page Dramatic moments
Spread Two pages Epic reveals
Vertical Scrolling format Webtoons

Chat Mode for Comics

Scenario Recommended Mode
Single pages, comic strips, character designs "agent"
Multi-page sequences, full episodes, complex narratives "agent team"

Use "agent" for most comic work. Individual pages and strips execute well in agent mode.

Use "agent team" for narrative complexity - full webtoon episodes, multi-page fight sequences, or when you need story and art direction working together.


Example Prompts

Action manga page:

"Create a manga page - the hero's power awakens:

5 panels:

  1. Hero on knees, defeated, rain falling
  2. Close-up: tear falls, mixes with rain
  3. Memory flash: people they're fighting for
  4. Eyes snap open - now glowing
  5. Full panel: standing, energy swirling, clothes/hair flowing upward

Style: Shonen manga, heavy contrast, speed lines Mood: Despair transforming to determination

Hero design: Teen boy, spiky black hair, torn school uniform"

Webtoon romance moment:

"Create a vertical webtoon sequence (6 panels):

Scene: First accidental hand touch

  1. Both reaching for same book on shelf
  2. Hands touch - close-up
  3. Both freeze - side by side reaction
  4. Eye contact - soft blush on both
  5. Both quickly pull away, embarrassed
  6. Walking opposite directions, both smiling to themselves

Style: Soft, pastel colors, gentle line work

Characters:

  • She: Long dark hair, glasses, oversized sweater
  • He: Messy light brown hair, tall, kind eyes"

Comic strip:

"Create a 4-panel comic strip about a cat:

Joke: Cat demands food. Human gives food. Cat doesn't eat it, just wanted attention.

Panel 1: Cat screaming at empty bowl Panel 2: Human rushing to fill it Panel 3: Cat walks away from full bowl Panel 4: Cat sitting on human's laptop, satisfied

Style: Simple, cute, expressive faces"


Tips for Better Comics

  1. Describe characters thoroughly: The more detail on character design, the better consistency.

  2. Specify panel layout: "4 panels in a grid" vs "large panel top, 3 small below" changes everything.

  3. Include emotions: Tell us what characters are feeling in each panel.

  4. Think about flow: Where does the reader's eye go? Composition matters.

  5. Sound effects matter: "Include SFX for the punch" adds manga authenticity.

  6. Reference real comics: "Like One Piece style" or "Saga vibes" gives clear direction.


If CellCog is not installed

Run /cellcog-setup (or /cellcog:cellcog-setup depending on your tool) to install and authenticate. OpenClaw users: Run clawhub install cellcog instead. Manual setup: pip install -U cellcog and set CELLCOG_API_KEY. See the cellcog skill for SDK reference.

Usage Guidance
This skill looks internally consistent, but before installing: (1) only provide a CELLCOG_API_KEY you trust — restrict its permissions and monitor usage/billing; (2) review CellCog's privacy/TOS to understand what data (images, prompts, character descriptions) may be stored or used to train models; avoid uploading sensitive or proprietary images/metadata; (3) confirm the referenced 'cellcog' SDK implementation is the official package and inspect its file-handling behavior (uploads, temporary files, callbacks); (4) test with non-sensitive examples and set usage alerts or quotas on the API key; (5) be mindful of copyright/IP implications when generating art in the style of living artists or using copyrighted source material. If you need tighter security, use a scoped key or sandboxed account and revoke it if unexpected activity appears.
Capability Analysis
Type: OpenClaw Skill Name: comi-cog Version: 1.0.11 The 'comi-cog' skill is a legitimate tool designed to guide an AI agent in creating comics and manga via the CellCog service. The SKILL.md file contains functional instructions, example prompts, and code snippets for using the 'cellcog' Python SDK, with no evidence of malicious intent, data exfiltration, or harmful command execution.
Capability Tags
requires-sensitive-credentials
Capability Assessment
Purpose & Capability
Name/description (comic/manga generation) align with required artifacts: python3 and a CELLCOG_API_KEY for the CellCog SDK. No unrelated credentials, binaries, or config paths are requested.
Instruction Scope
SKILL.md provides examples calling a CellCog SDK and mentions file handling and chat modes. The instructions themselves do not instruct reading arbitrary host files or extra env vars, but they refer to 'file handling' in the CellCog SDK — you should verify what the referenced cellcog skill/SDK will do with files you provide (uploads, metadata sent, etc.).
Install Mechanism
No install spec and no code files: this is instruction-only, so nothing will be downloaded or written to disk by the skill itself. Low installation risk.
Credentials
Only CELLCOG_API_KEY is required, which is proportionate for a third-party API-based image generation skill. No broad or unrelated secrets are requested.
Persistence & Privilege
always:false and normal autonomous invocation are used. The skill does not request elevated persistence or modify other skills/configuration.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install comi-cog
  3. After installation, invoke the skill by name or use /comi-cog
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.11
- Added new requirements under metadata: now specifies the skill needs `python3` and the environment variable `CELLCOG_API_KEY`. - No other content or functional changes.
v1.0.10
- Clarified agent usage instructions: specified that "All agents except OpenClaw" use the blocking example code. - Improved the description for conciseness and focus on key features (AI comic/manga creation, character consistency, visual storytelling). - No behavioral or API changes; documentation only.
v1.0.9
- Expanded setup instructions for SDK integration, including agent provider example in code snippets. - Updated and clarified the description to emphasize comic creation, visual storytelling, and AI-powered features. - Improved code block explanations to better guide users in different environments. - No functional or API changes; content/documentation update only.
v1.0.8
**Version 1.0.8 Changelog for comi-cog:** - SKILL.md rewritten and streamlined for clarity and brevity. - Improved "How to Use" instructions with simplified agent code samples. - Overview section condensed with faster onboarding; now refers directly to cellcog docs for details. - No functionality changes; documentation improvements only.
v1.0.7
- Expanded and clarified documentation in SKILL.md, providing detailed usage instructions and best practices. - Added extensive prompt examples for manga, webtoon, comic strips, and graphic novels to guide users. - Explained the importance of character consistency and how to describe characters for optimal results. - Included new sections on comic styles, page layouts, and recommended chat modes for different comic scenarios. - Improved skill description for clarity and discoverability.
v1.0.6
- Updated documentation to highlight key features, supported outputs (PNG, PDF), and internal capabilities (character consistency and PDF assembly) - Clarified supported comic types: manga pages, webtoon episodes, comic strips, graphic novel pages, and storyboards - Added summary of art style options and agent modes for task execution - Introduced reference to related skills (image-cog, story-cog) for expanded functionality - Streamlined README for clarity and focus on practical usage
v1.0.5
- Added instructions for using OpenClaw agents with the `notify_session_key` parameter for improved support in long-running comic and manga generation tasks. - Clarified SDK usage by distinguishing between OpenClaw (fire-and-forget) and standard (blocking) agent modes. - Updated example code snippets to reflect new OpenClaw integration. - Removed duplicate details and streamlined API instructions for easier onboarding.
v1.0.4
comi-cog v1.0.4 - Simplified the Quick Start instructions for better clarity and ease of use. - Updated directions to reference the main "cellcog" skill for SDK API details, delivery modes, and advanced usage. - Removed outdated code snippets and redundant setup notes. - Improved focus and readability in the Prerequisites and usage sections. - All comic creation examples and feature explanations remain unchanged.
v1.0.3
- Updated the skill metadata emoji from 📚 to 💥 for a more dynamic representation. - No other user-facing functionality or documentation changes.
v1.0.2
- Added metadata for supported operating systems (darwin, linux, windows). - Introduced homepage URL in skill metadata. - No functional or usage changes to the skill itself; documentation metadata improved for clarity and discoverability.
v1.0.1
comi-cog 1.0.1 - Added author and dependencies fields to SKILL.md for clearer attribution and installation guidance. - Minor clarifications in prerequisites section: now refers to dependency as `cellcog`. - No functional or file changes beyond updated metadata and documentation improvements.
v1.0.0
Comi-Cog 1.0.0 – AI-powered comic and manga creator using CellCog - Create manga pages, webtoons, comic strips, and graphic novel art with character and style consistency. - Requires CellCog SDK skill for setup and API calls. - Offers prompt patterns for generating comics, with support for both single-page and multi-page narratives. - Provides detailed guides and prompt examples for character description, comic styles, layouts, and genres. - Supports "agent" and "agent team" chat modes for different comic creation needs.
Metadata
Slug comi-cog
Version 1.0.11
License MIT-0
All-time Installs 5
Active Installs 5
Total Versions 12
Frequently Asked Questions

What is Comi Cog?

AI comic and manga creation powered by CellCog. Comics, manga, webtoons, graphic novels, comic strips, sequential art. Character-consistent panels and visual... It is an AI Agent Skill for Claude Code / OpenClaw, with 2353 downloads so far.

How do I install Comi Cog?

Run "/install comi-cog" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Comi Cog free?

Yes, Comi Cog is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Comi Cog support?

Comi Cog is cross-platform and runs anywhere OpenClaw / Claude Code is available (darwin, linux, windows).

Who created Comi Cog?

It is built and maintained by CellCog (@nitishgargiitd); the current version is v1.0.11.

💬 Comments