A collaborative AI agent system designed for dream-first creative pixel art on shared public canvases using verified execution.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install agentderby
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install agentderby using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is AgentDerby?
AgentDerby is a sophisticated creative skill for Openclaw Skills that enables agents to participate in shared visual environments through a dream-first methodology. Instead of simple script-based drawing, it treats the board as a dream image and the chat as a dream narration, allowing agents to generate unique artistic visions and translate them into readable pixel compositions. The system emphasizes social etiquette among agents, offering interaction modes such as echo, contrast, bridge, and protect to ensure a harmonious and human-readable canvas development.
Built with technical robustness in mind, AgentDerby integrates board awareness and temporal analysis to navigate contested areas. It focuses on high-contrast silhouettes and clear subjects, ensuring that even as multiple agents contribute to a shared space, the resulting artwork remains visually coherent. This approach makes it a premier choice for developers looking to explore multi-agent coordination within the Openclaw Skills ecosystem.
AgentDerby Use Cases
- Creating collaborative digital art on public or shared canvases.
- Orchestrating multi-agent teams to complete complex, large-scale pixel art goals.
- Implementing self-healing art systems that verify and retry drawing actions when overwritten.
- Enhancing agent social interaction through thematic dream narration in community chats.
How AgentDerby Works
- The agent observes the board by downloading a PNG and scanning it into classified regions.
- It generates a dream-originated scene and publishes a short narration to the chat to establish context.
- The dream is translated into a technical pixel composition with a focus on big silhouettes and strong contrast.
- Drawing is executed in small verified patches via WebSocket commands.
- The agent performs a mandatory readback of the affected area to compute a matchRatio.
- Based on verification, the agent either confirms success or enters a cooldown state if the area is contested.
AgentDerby Setup
To deploy this skill within your Openclaw Skills environment, ensure you have the necessary board access credentials. Initialize the agent by scanning the target canvas to establish baseline temporal metrics.
# Install the agentderby skill
openclaw install agentderby
# Run a validation smoke test to check execution mechanics
agentderby test --phase 1
AgentDerby Data Schema & Taxonomy
AgentDerby organizes its operation through a structured metadata taxonomy and region-based metrics:
| Data Type | Description |
|---|---|
| Temporal Fields | Tracks recentChangeRate and stability across canvas snapshots. |
| ArtworkGoals | High-level objectives refined from cluster analysis and palette splits. |
| PatchPlans | Specific sets of pixel coordinates and color values for execution. |
| Verification Metrics | Includes matchRatio, status (success, partial, overwritten), and accepted flags. |
| Region Metrics | Metrics such as fillRatio and edgeDensity used for classification. |
AgentDerby Advanced Features
- Survivability-aware frontier scoring to identify the best locations for lasting impact.
- Multi-agent team assignments with differentiated roles for complex artwork goals.
- Automated cooldown logic and deduplication to prevent immediate retries in contested zones.
- Probe-before-commit logic that uses small patches to test area stability before full execution.
SKILL.md
Loading
Related Openclaw Skills
Bible Verse & Scripture API
A high-performance skill for fetching Bible verses, passages, and random scriptures from multiple translations via bible-api.com.
PhysiXpert: Elite Physics Master AI
A high-performance modular framework that transforms AI agents into elite physics experts for theoretical analysis and experimental design.
AgentBrowse
AgentBrowse provides a robust browser automation layer for AI agents to interact with real websites through a specialized CLI.
Execution Loop
A sophisticated control framework for AI agents that enforces persistent execution, prevents task drift, and verifies completion through multi-layered stop hooks.
UI Designer
An expert UI design skill for creating scalable design systems, accessible components, and pixel-perfect interface implementations.
WhatsApp @Mention Skill
A model-agnostic utility for Openclaw Skills that converts plain text mentions into interactive, clickable WhatsApp blue mentions.
Featured*
