Sticker Cog

AI sticker pack generation powered by CellCog. Custom stickers, emoji sets, WhatsApp stickers, Telegram stickers, Discord emoji, Slack reactions. Character-c...

Security Scan

The skill's requirements and instructions are consistent with a CellCog-powered sticker generator, but it will send images and prompts to an external service and the SKILL.md references a SDK dependency without an install step — review privacy and runtime setup before installing.

Purpose & Capability

Name/description match the observed requirements: python3 and a CELLCOG_API_KEY are reasonable for a CellCog SDK-based sticker generator. The SKILL.md declares a dependency on a 'cellcog' package but provides no install spec (no pip/installer instructions), which is a minor mismatch — the skill assumes the runtime has the SDK available.

Instruction Scope

The instructions focus on calling the CellCog service via its SDK. They explicitly reference uploading user photos and generating/storing platform-ready assets, which means user images and prompts will be transmitted to CellCog's servers. The instructions do not appear to read unrelated system files or request unrelated environment variables, but they do point agents to another 'cellcog' skill for full SDK reference.

Install Mechanism

No install spec is present (instruction-only), so nothing is written to disk by the skill itself — this is low risk. The only minor issue is the SKILL.md's declared dependency ('cellcog') with no guidance for installing it; the runtime must already provide that SDK or the agent will need to install it.

Credentials

Only a single environment variable (CELLCOG_API_KEY) is required, which is proportional to a cloud API-based image generation service. No unrelated credentials or broad filesystem/config path access are requested.

Persistence & Privilege

The skill is not set always:true and uses normal agent invocation. It does not request persistent system-wide privileges in the manifest. Autonomous invocation is allowed (platform default) but is not combined with other high-risk flags here.

Assessment

This skill appears to do what it says: it calls CellCog to generate sticker packs and needs your CELLCOG_API_KEY and Python/SDK support. Before installing: 1) Be aware that any photos or prompts you provide will be sent to CellCog's servers — avoid uploading sensitive personal or confidential images. 2) Confirm you trust CellCog's privacy and data retention policies (see https://cellcog.ai). 3) The SKILL.md names a 'cellcog' dependency but gives no install steps; ensure your environment provides that SDK (or be prepared to let the agent/pip install it). 4) If you prefer tighter control, do not provide the CELLCOG_API_KEY to the agent or restrict when the skill may be invoked. If you want more certainty about runtime behavior, request the full SKILL.md/SDK reference or an install spec from the publisher.

Sticker Cog - One Character, Twenty Expressions, Perfect Consistency

One character. Twenty expressions. Every sticker looks like the same person. That's the hard part — and that's what CellCog does better than anyone.

Generating one cute sticker is easy. Every AI tool can do that. But generating a complete pack of 20 stickers where the character has the same face, same style, same proportions — just different expressions and poses? That's consistency at scale, and it's where most AI tools completely fall apart. By sticker #10, your character looks like a different person.

CellCog maintains character identity across every single generation, delivers with transparent backgrounds, and packages everything for your target platform — WhatsApp, Telegram, Discord, Slack, or iMessage.

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 Sticker Packs Are Hard

The challenge isn't making one sticker. It's making twenty that look like they belong together.

The Problem	Why It's Hard	How CellCog Solves It
Character drift	Each AI generation drifts slightly — different face shape, different proportions	CellCog's reference image threading maintains identity across all generations
Style consistency	Line weight, color palette, and art style shift between generations	The agent establishes a visual style guide before generating and enforces it throughout
Expression range	Getting 20 genuinely different expressions without changing the character	CellCog plans the full expression set upfront, then generates each with the character locked
Transparent backgrounds	Many AI models struggle with clean transparency	Dedicated transparent background model produces clean cutouts
Platform formatting	Each messaging platform has different size, format, and packaging requirements	Automatic conversion and optimization for WhatsApp, Telegram, Discord, Slack, iMessage

---

What Sticker Packs You Can Create

Character Sticker Packs

Design a character and get a complete expression set:

• Original Characters: "Create a sticker pack with a friendly fox character — 20 expressions for everyday messaging"
• Chibi/Anime Style: "Design a chibi version of a warrior princess with 15 battle and everyday expressions"
• Mascot Packs: "Create a sticker pack for our startup mascot — a robot named Bolt"
• Couple/Duo Packs: "Design two best friend characters and create stickers of them together"

Example prompt:

"Create a WhatsApp sticker pack — 20 stickers:

Character: A chubby orange tabby cat named Mochi Style: Kawaii, round, simple clean lines, warm color palette

Expressions:

1. Happy/smiling
2. Laughing (rolling on back)
3. Crying (dramatic anime tears)
4. Angry (puffed up)
5. Sleeping (curled up, zzz)
6. Eating (bowl of ramen)
7. Love (heart eyes)
8. Confused (head tilt, question mark)
9. Excited (jumping)
10. Thumbs up
11. Facepalm
12. Waving hello
13. Sad (rain cloud above)
14. Celebrating (party hat, confetti)
15. Thinking (chin on paw)
16. Shocked/surprised
17. Cool (sunglasses)
18. Sick (thermometer, blanket)
19. Working (tiny laptop)
20. Goodnight (moon, stars)

Every sticker MUST look like the same cat. Format for WhatsApp: 512×512, WebP, transparent backgrounds, under 100KB each."

Brand & Team Emoji

Custom reactions for your workspace:

• Company Mascot Emoji: "Turn our company mascot into a Slack emoji set with 15 work-related reactions"
• Team Reactions: "Create a custom emoji set for our engineering Slack — ship it 🚢, LGTM, hotfix, etc."
• Themed Packs: "Make a set of 10 food-themed reaction stickers for our restaurant's Discord"

Example prompt:

"Create a Slack emoji set for our engineering team:

Style: Minimal, flat design, consistent with tech aesthetic

Reactions needed:

• 'Ship it' (rocket launching)
• 'LGTM' (green checkmark with sparkles)
• 'Hotfix' (fire + wrench)
• 'In review' (magnifying glass)
• 'Blocked' (stop sign)
• 'Needs coffee' (coffee cup)
• 'On call' (phone ringing)
• 'Deployed' (party popper)
• 'Rollback' (reverse arrow)
• 'Weekend mode' (palm tree)

All 128×128 PNG with transparent backgrounds. Keep them readable at small sizes."

From Reference Photos

Turn real photos into sticker packs:

• Pet Stickers: "Turn this photo of my golden retriever into a 15-sticker cartoon pack"
• Self-Portrait Stickers: "Create a cartoon sticker pack based on my photo — keep the likeness but make it cute"
• Group Stickers: "Use this team photo to create cartoon versions of us for our company Slack"

Example prompt:

"Here's a photo of my cat Luna: [upload photo]

Create a Telegram sticker pack (15 stickers) based on Luna's actual appearance:

• Keep her black fur, green eyes, and white chest patch
• Style: Cute cartoon, slightly exaggerated proportions (bigger head, bigger eyes)

Expressions: happy, sleepy, hungry, judgmental, playful, zoomies, loaf position, stretching, knocking something off a table, hiding in a box, batting at a toy, yawning, grooming, startled, and one with a crown (queen energy)

Every sticker must be recognizably Luna. 512×512 PNG, transparent backgrounds."

Professional Reaction Stickers

Workplace-appropriate reactions:

• Project Status: "Create stickers for project management — Approved, In Progress, Needs Revision, Shipped, Blocked"
• Meeting Reactions: "Make stickers for virtual meetings — Agree, Disagree, Question, Great Idea, Let's Table This"
• Feedback Stickers: "Create a set of code review reaction stickers — LGTM, Nit, Bug, Refactor Needed, Nice Pattern"

Platform-Specific Packs

Optimized for where your audience lives:

• WhatsApp Packs: "Create a complete WhatsApp sticker pack — 30 stickers, ready to import"
• Telegram Sets: "Build a Telegram sticker set with 25 stickers"
• Discord Emoji: "Create 20 custom Discord server emoji"
• iMessage Stickers: "Design a sticker pack for iMessage with 15 expressions"

---

Sticker Styles

Style	Characteristics	Best For
Kawaii	Round, cute, simple, big eyes	Casual messaging, cute characters
Flat Design	Clean lines, solid colors, minimal shading	Professional/brand emoji
Cartoon	Expressive, dynamic, detailed	Character-driven packs
Pixel Art	Retro, nostalgic, grid-based	Gaming communities, nostalgia
Watercolor	Soft, artistic, painterly	Artistic/lifestyle brands
Line Art	Black outlines, minimal color	Elegant, universal
Chibi/Anime	Exaggerated proportions, big heads	Anime communities, expressive
3D Render	Glossy, dimensional, modern	Tech brands, premium feel

---

Platform Specifications

Platform	Format	Size	Per Pack
WhatsApp	WebP, transparent	512×512, <100KB each	3-30 stickers
Telegram	PNG or WebP, transparent	512×512	Up to 120 stickers
Discord	PNG or GIF	128×128 or 320×320	Per slot
Slack	PNG, GIF, or JPG	128×128	Per slot
iMessage	PNG, transparent	300×300 recommended	10-40 stickers
General Use	PNG, transparent	Any size	Any count

Tell CellCog which platform you're targeting and it handles the formatting automatically — dimensions, file format, file size optimization, and transparent background generation.

---

Chat Mode for Sticker Packs

Scenario	Recommended Mode
Single sticker pack (10-20 stickers), emoji sets	"agent"
Multiple coordinated packs, brand sticker system, large sets (30+)	"agent team"

Use "agent" for most sticker work. A single pack with 10-20 stickers executes well in agent mode.

Use "agent team" for brand-level sticker systems — when you need multiple packs that share a design language, or very large sets where consistency across 30+ stickers benefits from deeper creative planning.

---

Example Prompts

Quick emoji set:

"Create 10 food-themed reaction emoji for our team's Slack:

Style: Cute, kawaii, each food item has a face Include: Pizza (excited), Sushi (cool), Taco (party), Coffee (sleepy), Burger (happy), Donut (love), Broccoli (disgusted), Ice cream (melting), Ramen (slurping), Cookie (waving)

128×128 PNG, transparent backgrounds."

Character pack from scratch:

"Design a character and create a full Telegram sticker set:

Character: A tiny astronaut with an oversized helmet, floating in space Personality: Curious, clumsy, optimistic Style: Flat illustration, limited color palette (white suit, blue visor, orange accents)

20 stickers covering common messaging scenarios: Hi, Bye, Thanks, Sorry, LOL, Crying, Love, Angry, Thinking, Sleeping, Celebrating, Confused, Excited, Thumbs up, Facepalm, Mind blown, Running late, On my way, Good morning, Good night

512×512 PNG, transparent backgrounds. The astronaut must look identical in every sticker."

Brand mascot stickers:

"Our company mascot is a friendly blue owl named Hoot. Here's the reference image: [upload mascot image]

Create a WhatsApp sticker pack (20 stickers) using Hoot:

• Keep Hoot's exact design, colors, and proportions
• Add work-related expressions: 'Great job!', 'Deadline approaching', 'Meeting time', 'Need coffee', 'On it!', 'Thinking...', 'Approved!', 'Let's discuss', etc.
• Mix of work reactions and casual expressions

Format: 512×512 WebP, transparent backgrounds, under 100KB each. Also create a set of 10 as 128×128 PNG for Slack emoji."

Pet sticker pack:

"Create a sticker pack from my dog's photo: [upload dog photo]

My dog is a corgi named Biscuit. Turn Biscuit into a cartoon character (keep the corgi proportions, stumpy legs, big butt) and create 15 stickers:

Happy wiggle, Zoomies, Sploot, Begging for treats, Judging you, Sleeping in funny position, Barking, Head tilt, Catching a ball, Rolling in grass, Wet dog, Wearing a tiny hat, Digging, Tail wagging, Puppy eyes

For Telegram. 512×512, transparent."

---

Tips for Better Sticker Packs

1. Describe the character in detail: "Orange cat" is vague. "Chubby orange tabby with green eyes, white chest patch, round face, always slightly smug expression" gives CellCog what it needs for consistency.

2. Plan your expressions upfront: List all the expressions you want before generating. This helps CellCog plan the full set and avoid drift.

3. Specify the style clearly: "Kawaii", "flat design", "chibi anime" — one clear style direction keeps the whole pack cohesive.

4. Include reference images when possible: If you're making stickers from a pet photo, mascot design, or existing character, upload the reference. CellCog uses it to maintain likeness.

5. Think about readability at small sizes: Stickers are viewed tiny. Bold outlines, simple shapes, and clear expressions read better than detailed illustrations.

6. Platform first: Different platforms have wildly different requirements. "For WhatsApp" vs. "For Discord" changes dimensions, format, and file size targets. Always specify.

---

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.