A sophisticated routing and governance engine that manages AI agent skill selection through a layered taxonomy and proactive risk assessment.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install skill-taxonomy-router-pro
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 skill-taxonomy-router-pro 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 Taxonomy Router Pro?
Skill Taxonomy Router Pro is a high-performance management layer designed to govern how AI agents interact with their available capabilities. By implementing a structured domain taxonomy and a multi-level risk model, it ensures that only the most relevant and safe skills are loaded for any given task. This prevents context bloat and mitigates system risks by adhering to a minimum-necessary-loading strategy.
Built for power users of Openclaw Skills, this tool provides a comprehensive framework for classifying new skills, detecting functional overlaps, and maintaining a persistent routing policy. It transforms a chaotic directory of scripts into an organized, searchable, and secure command center, allowing agents to navigate complex workflows with surgical precision.
Skill Taxonomy Router Pro Use Cases
- Orchestrating multi-agent workflows where distinct roles like retrieval and publishing are required.
- Governing system security by assessing risk levels (R0-R4) before executing action-capable skills.
- Managing large skill libraries by classifying them into domains such as Development, Infrastructure, or Finance.
- Automating the intake and review process for newly downloaded Openclaw Skills to prevent redundant functionality.
- Optimizing LLM context windows by loading only the single most specific skill needed for a task.
How Skill Taxonomy Router Pro Works
- Intent Inference: The router analyzes the user request to determine the primary object and required action.
- Taxonomy Matching: It searches the hierarchical domain taxonomy to find the most specific matching subdomain.
- Risk Evaluation: The system assesses potential threats across dimensions like data egress and local system impact to ensure the safest path.
- Skill Selection: Based on the routing algorithm, it selects the highest specificity skill with the lowest context cost.
- Post-Action Logging: Every routing decision and usage metric is logged to refine future selection accuracy and track skill frequency.
Skill Taxonomy Router Pro Setup
To integrate this router into your environment, initialize the index and set up the tracking logs:
# Rebuild the installed-skill index
python3 scripts/update_skill_index.py --log
# Intake a new skill for review
python3 scripts/intake_skill.py <skill-name>
# Review high-frequency skill usage
python3 scripts/track_skill_usage.py report --limit 30
Skill Taxonomy Router Pro Data Schema & Taxonomy
The skill maintains a structured metadata repository within the references/ directory to manage the lifecycle of Openclaw Skills:
| File | Purpose |
|---|---|
skill-index.md |
The master registry of all installed and classified skills. |
usage-stats.json |
Quantitative data on skill invocation frequency. |
routing-decisions.jsonl |
An append-only log of intent-to-skill mapping history. |
skill-overlap-map.json |
A taxonomy of canonical skills versus their functional variants. |
backlog-priority.md |
A prioritized list of unclassified skills awaiting formal review. |
Skill Taxonomy Router Pro Advanced Features
- Risk-Aware Orchestration: Automatically blocks high-risk operations in unreviewed skills by default.
- Overlap Fusion: Identifies duplicate capabilities and generates a canonical-first routing view to reduce redundancy.
- Automated Intake Workflow: A dedicated inbox system for sandboxing and screening new Openclaw Skills before they enter the production index.
- Taxonomy Expansion Logic: Intelligent detection of backlog clusters that suggest the need for new subdomains.
- Multi-Skill Composition: Ruleset for combining complementary skills, such as pairing a CRM tool with an outbound messaging automation.
SKILL.md
Loading
Related Openclaw Skills
Writing Agent Creator
An automated system for generating and managing persistent, specialized AI writing agent configurations for the OpenClaw ecosystem.
Qwen
A comprehensive framework for routing and debugging Qwen model workflows across cloud-hosted and local environments.
MuAPI Platform Utilities
A set of utility scripts for configuring muapi.ai API keys and managing asynchronous generation requests.
Copywriting Specialist
A professional direct-response copywriting engine that transforms products into sales using industry-proven psychological frameworks.
Recipe: Email Google Drive Link
A productivity recipe that automates the process of locating a Google Drive file, granting permissions, and emailing the link via Gmail.
Requirement Expander (req-expander)
A specialized tool designed to expand brief requirement descriptions into comprehensive technical design points and testing specifications.
Featured*
