A comprehensive master router and cognitive engine that unifies disparate domains into a single, intelligent interface for AI agents.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install omni-master
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 omni-master 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 OMNI?
OMNI serves as the central nervous system for your agentic environment, providing a unified interface that spans across every capability domain from coding and DevOps to finance and smart home management. As a cornerstone of the Openclaw Skills ecosystem, OMNI acts as a master router that intelligently classifies user tasks and loads specific reference protocols on demand. This progressive disclosure approach ensures that the agent remains lean and focused while maintaining access to a vast library of technical knowledge and execution strategies.
At its core, OMNI implements a sophisticated cognitive architecture known as the Brain Protocol. This protocol enforces rigorous logic, error memory, and anti-hallucination checks before any domain-specific action is taken. Whether you are building complex CI/CD pipelines, automating financial reports, or managing IoT devices, OMNI provides the necessary glue to connect specialized tools and ensure reliable, high-precision results.
OMNI Use Cases
- Orchestrating multi-domain workflows such as scraping data, analyzing it, and then generating a visual report.
- Implementing a cognitive safety layer to prevent repetitive errors and hallucinations in AI agent sessions.
- Routing complex or vague user requests to the appropriate specialized Openclaw Skills or reference protocols.
- Managing full-stack development cycles including code generation, security auditing, and cloud deployment.
How OMNI Works
- The Brain Protocol is initialized to load cognitive engine rules and check the persistent mistake log to avoid repeating past errors.
- The input task is analyzed for signal words to determine which specific domains (e.g., Coding, Data, Security) are required.
- Relevant reference files are loaded dynamically to provide the agent with deep domain knowledge without cluttering the context window.
- A detailed execution plan is formulated, identifying the best available CLI tools or specialized skills to use.
- The agent executes the plan and verifies the output quality against defined performance and QA standards.
- Success patterns and errors are logged to the memory system to optimize future interactions within the Openclaw Skills framework.
OMNI Setup
To integrate OMNI into your agent environment, ensure the directory structure is preserved so the router can access its reference library.
# Navigate to your skills directory
cd path/to/your/agent/skills
# Clone or move the omni skill directory
mv /downloads/omni .
# Verify the architecture is intact
ls -R omni/
# Ensure the brain protocol is accessible
cat omni/references/brain.md
OMNI Data Schema & Taxonomy
OMNI organizes its intelligence through a structured hierarchy of references and persistent memory logs to ensure high-fidelity operations within Openclaw Skills.
| Component | Type | Description |
|---|---|---|
references/brain.md |
Protocol | The primary cognitive engine and logic ruleset. |
memory/mistakes.json |
Log | A persistent JSON record of failed approaches to ensure loop detection. |
memory/patterns.json |
Log | Storage for successful, reusable workflows and logic patterns. |
references/*.md |
Domain | Deep-dive documentation for specific domains like DevOps, Finance, or Coding. |
scripts/router.sh |
Script | Internal logic for routing tasks between reference files. |
OMNI Advanced Features
- Advanced Loop Detection: Automatically identifies and halts failing command sequences to pivot to new strategies.
- Multi-Agent Orchestration: Capabilities to spawn, collaborate with, and manage parallel agent processes.
- Precision Memory Management: Logs every decision and error to create a self-improving local knowledge base.
- Progressive Context Disclosure: Maintains a lean context window by loading only the most relevant domain data for the current task.
- Cross-Skill Synergy: Acts as a management layer for other Openclaw Skills, filling gaps between specialized tools.
SKILL.md
Loading
Related Openclaw Skills
iaskaster
iaskaster is a comprehensive Bazi fortune-telling and destiny analysis skill that generates professional PDF reports based on traditional Chinese metaphysics.
Life Logging Analyzer
A personal data synthesis engine that transforms fragmented records into structured life narratives and actionable insights.
Go Pro: Advanced Go 1.21+ Development
An expert-level Go development skill for building high-performance, concurrent, and production-ready microservices using Go 1.21+ features.
Alibaba Cloud Skill Creator
A specialized workflow tool for standardizing the creation, migration, and benchmarking of AI agent skills within the alicloud-skills repository.
skill-evolver
An autonomous evolution engine that transforms raw AI execution data into optimized, production-ready skills through feedback loops and pattern extraction.
Aviation Weather (L402)
A professional-grade aviation weather tool providing METAR observations, TAF forecasts, and station discovery via the L402 API.
Featured*
