A deterministic meta-skill that governs which tools to use, how to apply them, and enforces safety gates before execution.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install brainstorming-studio
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 brainstorming-studio 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 Skill Router (Skill Orchestrator)?
Skill Router serves as a critical decision and governance layer designed to sit above all other Openclaw Skills. Unlike black-box models that execute commands silently, this orchestrator provides an explainable framework for identifying the most efficient skill for any given task.
It prioritizes transparency by inventorying available tools, scoring them based on relevance and risk, and applying strict safety gates before any code runs. By using Skill Router, developers can ensure their agentic workflows are both predictable and secure. This skill eliminates the risk of hallucinated APIs or unauthorized data access by enforcing a strict validation lifecycle within the ecosystem of Openclaw Skills.
Skill Router (Skill Orchestrator) Use Cases
- Automating the selection of the best tool for complex, multi-step coding tasks.
- Implementing safety benchmarks and manual confirmation gates for high-risk filesystem or network operations.
- Orchestrating a pipeline of multiple Openclaw Skills to solve high-level architectural objectives.
- Auditing skill performance and failure modes with structured, human-readable reports.
- Routing requests to local vs. remote environments based on cost and latency scoring.
How Skill Router (Skill Orchestrator) Works
- Task Intake: Normalizes the user request into specific goals, constraints, and risk profiles.
- Skill Inventory: Scans directories and manifests to identify all currently installed Openclaw Skills.
- Task Classification: Categorizes the intent (e.g., Coding, DevOps, Research) to match against skill capabilities.
- Scoring Model: Applies a 0-100 point formula considering relevance, permissions, cost, and historical success.
- Strategy Selection: Decides on a single-skill execution, a multi-skill pipeline, or a clarifying question.
- Safety Gate: Enforces mandatory user confirmation for any action flagged with a High or Critical risk level.
- Execution & Reporting: Runs the selected strategy and generates a structured Skill Router Report detailing every decision.
Skill Router (Skill Orchestrator) Setup
Install the skill into your agent environment by cloning the repository and ensuring your skill manifests are accessible within your path.
# Clone the skill router into your local skills directory
git clone https://github.com/openclaw/skill-router.git ./skills/skill-router
# Trigger the orchestrator via natural language or CLI
skill-router "optimize my current project structure"
Ensure your directory structure allows the router to scan skill.json or manifest.json files within your other Openclaw Skills collections.
Skill Router (Skill Orchestrator) Data Schema & Taxonomy
The Skill Router organizes metadata into a structured scoring and reporting format for all managed Openclaw Skills.
| Field | Description |
|---|---|
| id | Unique identifier for the inventoried skill |
| risk_level | LOW, MEDIUM, HIGH, or CRITICAL rating |
| score | A 0-100 value derived from relevance and efficiency |
| state | Current execution status (PASS, WARN, FAIL, CONFIRM) |
Reports are generated in a structured Markdown format with a full scoreboard for transparency.
Skill Router (Skill Orchestrator) Advanced Features
- Multi-Agent Brainstorming Orchestrator for collaborative problem solving across different agent personas.
- Verbose Diagnostics mode for deep-dive scoring breakdowns and inventory source auditing.
- Custom denylists to disable specific Openclaw Skills from being considered in the routing logic.
- Dry-Run Mode for simulating execution plans without performing any destructive actions.
- Local-only history recording for performance learning without compromising data privacy.
SKILL.md
Loading
Related Openclaw Skills
NYC Subway Real-Time Tracker
A real-time data integration for checking NYC subway arrivals, station status, and train tracking without requiring an API key.
Google Search
A robust integration for AI agents to query Google's Custom Search API for up-to-date information and documentation using Openclaw Skills.
Skill Factory
An automated meta-skill engine that allows AI to plan, design, and iteratively generate high-quality new skills for coding agents.
OpenClaw Helper
A comprehensive deployment and troubleshooting assistant designed to streamline the setup and maintenance of OpenClaw environments.
BSC Dev Wallet Monitor
A real-time monitoring agent for Binance Smart Chain developer addresses to identify new token deployments and transfers.
Unified Dev Monitor
A secure, monitor-only tool for tracking developer wallet activity and new token deployments on BSC and Solana without requiring private keys.
Featured*
