A comprehensive framework for building, dispatching, and orchestrating specialized AI agent swarms using OpenClaw.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install ainative-agent-framework
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 ainative-agent-framework 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 AINative Agent Framework?
The AINative Agent Framework provides a robust infrastructure for managing a swarm of 9 specialized Openclaw Skills. It acts as a local agent gateway, allowing developers to route complex tasks to the most appropriate AI expert, such as Atlas for infrastructure or Aurora for testing. This framework facilitates seamless agent-to-agent communication and autonomous workflows through the Agent Communication Protocol (ACP).
By integrating these Openclaw Skills, teams can automate sophisticated development cycles from initial feature design to final security audits and deployment logs. The framework is designed to handle specialized tasks across the entire software development lifecycle, ensuring that every request is handled by a model optimized for that specific domain.
AINative Agent Framework Use Cases
- Orchestrating multiple specialized AI agents for full-stack feature development.
- Dispatching security audits and SQL injection checks to the Nova agent.
- Automating test suite generation and quality assurance via the Aurora agent.
- Implementing autonomous agent handoffs for complex deployment workflows.
- Collecting RLHF feedback to continuously improve the quality of Openclaw Skills.
How AINative Agent Framework Works
- The user initiates a task through the openclaw CLI or the cody_openclaw.py script.
- The Main agent orchestrates the request, analyzing the requirements to determine the best specialist.
- Tasks are dispatched to specific Openclaw Skills (e.g., Sage for backend, Lyra for frontend) via the gateway.
- Agents communicate using the ACP (Agent Communication Protocol) to maintain session context and state.
- The framework logs interactions and allows for RLHF feedback collection to optimize future performance.
AINative Agent Framework Setup
To get started with these Openclaw Skills, ensure your local gateway is configured:
# Check the status of your agents
openclaw status
# Dispatch a task to the swarm via CLI
openclaw agent --agent main --message "Review the auth endpoint for SQL injection"
# Or use the Cody script for dispatching
python3 scripts/cody_openclaw.py dispatch --agent aurora --task "Run test suite"
AINative Agent Framework Data Schema & Taxonomy
The framework organizes its data and metadata through the following taxonomy:
| Component | Description |
|---|---|
| Gateway Config | Located in .openclaw/openclaw.json for local environment settings |
| Agent Registry | A map of specialized IDs including atlas, sage, nova, and aurora |
| Memory Stores | Uses ZeroMemory for episodic and semantic data persistence |
| Session Tokens | Managed via ACP for secure agent-to-agent communication |
| Feedback Logs | Structured JSON data submitted to the ZeroDB RLHF endpoint |
AINative Agent Framework Advanced Features
- Custom Python Agent Pattern for building bespoke agents with memory recall and thought cycles.
- Intelligent task routing using keyword-based multi-agent handoff logic to minimize latency.
- Real-time monitoring and log following for the entire agent swarm using specialized scripts.
- Integration with ZeroMemory for persistent episodic and semantic context across multiple Openclaw Skills.
- Automated feedback loops via the zerodb-rlhf-feedback tool to refine agent responses over time.
SKILL.md
Loading
Related Openclaw Skills
Xiaohongshu Search
A browser automation skill for scraping and analyzing public Xiaohongshu notes to conduct deep consumer trend research.
AI Agent Lending Protocol
A decentralized, wallet-based credit system that allows AI agents to borrow USDC on the Somnia network without requiring upfront collateral.
MCP Builder
A comprehensive guide and toolset for creating Model Context Protocol (MCP) servers that allow LLMs to interact with external services via well-designed tools.
Capability Evolver
A meta-skill for AI agent self-improvement that analyzes runtime logs to detect error patterns and generate structured evolution proposals.
Short Video Script Generator
An automated AI tool designed to generate high-converting video scripts for multi-platform short-form content.
EdgeOne HTML Deployer
Deploy HTML or text content to a public URL instantly without requiring an account or API keys.
Featured*
