A sophisticated tool discovery and registration system featuring token-based routing, granular permissions, and sub-agent whitelisting.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install tool-registry
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 tool-registry 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 Tool Registry?
The Tool Registry skill serves as the central nervous system for tool management within an AI agent ecosystem. It provides a robust framework for developers to register functions, scripts, and APIs as searchable tools. By implementing an intelligent token-matching algorithm, it allows agents to resolve the correct tool for any given task dynamically, making it an essential component for any advanced Openclaw Skills implementation.
This skill goes beyond simple lookups by incorporating a multi-layered security and routing model. It categorizes tools by permission levels—ranging from read-only to administrative access—and maps them to specific agent types like 'explore' or 'verification'. This ensures that agents operate within safe boundaries while maintaining high efficiency in task execution.
Tool Registry Use Cases
- Implementing dynamic tool discovery in multi-agent workflows.
- Enforcing role-based access control (RBAC) for dangerous shell commands.
- Filtering available functionality based on the current agent personality or task phase.
- Searching for relevant tools using natural language keywords and aliases.
How Tool Registry Works
- The system processes user input or agent requests through a tokenization layer.
- It matches extracted tokens against registered tool names, aliases, and keywords.
- A scoring algorithm ranks results, giving double weight to alias matches and half weight to description matches.
- The registry applies permission filters (e.g., read, write, danger) to exclude unauthorized tools.
- Sub-agent type filters further narrow the list to tools appropriate for the current context (e.g., 'plan' vs 'explore').
- The most relevant, authorized tool is returned or executed via the registry's unified API.
Tool Registry Setup
To integrate this into your Openclaw Skills project, ensure you have the core registry script available in your scripts directory.
# Standard file structure
# tool-registry/
# ├── SKILL.md
# └── scripts/tool-registry.mjs
Initialize the registry in your JavaScript environment:
const { ToolRegistry } = require('./scripts/tool-registry.mjs');
const registry = new ToolRegistry();
Tool Registry Data Schema & Taxonomy
The Tool Registry organizes data using a structured JavaScript object for each tool. Below is the metadata taxonomy used by Openclaw Skills:
| Property | Type | Description |
|---|---|---|
| name | string | The primary identifier for the tool |
| aliases | string[] | Short names or alternative triggers |
| permission | string | Security levels: read (1), write (2), danger (3), admin (4) |
| agentTypes | string[] | Allowed sub-agents (e.g., explore, plan, verification) |
| keywords | string[] | Search terms used for the matching algorithm |
| execute | function | The asynchronous logic triggered by the tool |
Tool Registry Advanced Features
- Weighted Search Algorithm: Intelligent scoring that prioritizes specific matches over broad descriptions.
- Granular Permission Filtering: Dynamic user-level or session-level permission checks before tool exposure.
- Agent-Specific Toolsets: Capability to define 'whitelists' for different agent archetypes to prevent logic errors.
- Composable Filters: Chain multiple logic gates (permissions, agent types, and exclusions) to create hyper-specific execution contexts.
SKILL.md
Loading
Related Openclaw Skills
Hook System
A powerful tool interception framework to inject custom logic before and after AI tool execution.
DeepRead Insurance Document Processing
DeepRead Insurance is a high-accuracy AI skill designed to extract structured data from complex insurance documents and redact sensitive PII.
Parasite SEO
A strategic framework for publishing content on high-authority platforms to leverage their domain strength for immediate search rankings and AI citations.
Session Compactor
An automated context management tool that prevents token overflows by summarizing long conversation histories into concise, structured snapshots.
Usage Tracker
A comprehensive token monitoring and budget management system designed to track AI agent consumption and operational costs.
Jike Digest
An automated content curation tool that extracts the latest high-value posts from Jike topics and transforms them into structured Markdown digests.
Featured*
