← Back to Skills Marketplace
wzk332266

Skill Taxonomy Router

by Stairer · GitHub ↗ · v1.5.0 · MIT-0
cross-platform ⚠ suspicious
315
Downloads
0
Stars
0
Active Installs
8
Versions
Install in OpenClaw
/install skill-taxonomy-router
Description
Route user tasks to the most relevant skills using a layered taxonomy, risk model, and minimum-necessary-loading strategy. Use when deciding which skill to l...
Usage Guidance
This skill is policy-level documentation for routing tasks to other skills and is internally consistent. Before enabling it: (1) confirm your agent enforces the SKILL.md guidance — especially explicit user approval before loading action-capable or newly downloaded skills; (2) test the router in a controlled environment to ensure it doesn’t get invoked indiscriminately for every user task; (3) verify there are no local/hidden extensions or separate packages that implement automation or logging not present in this published package (the changelog mentions heavier private extensions that are intentionally excluded from the public release). Otherwise the package is low-risk because it contains only routing rules and asks for no credentials or installs.
Capability Analysis
Type: OpenClaw Skill Name: skill-taxonomy-router Version: 1.5.0 The skill-taxonomy-router is a governance and orchestration tool that manages a library of other skills using a detailed risk model (R0–R4) and taxonomy. While its instructions focus on security and organization, it requires high-risk capabilities, including executing multiple local Python scripts (e.g., scripts/update_skill_index.py, scripts/intake_skill.py) and performing automated directory deletion on the user's Desktop (scripts/cleanup_download_inbox.py --yes). These behaviors are plausibly necessary for its stated purpose but represent a significant attack surface and potential for unintended data loss, warranting a suspicious classification under the provided criteria.
Capability Assessment
Purpose & Capability
The name/description (skill router / taxonomy-based routing) match the SKILL.md content: classification rules, risk baselines, threat dimensions, and routing algorithm. The skill does not request unrelated credentials, binaries, or config access.
Instruction Scope
Runtime instructions are policy and decision rules (how to infer intent, choose domains, prefer low-risk skills). They do not instruct reading system files, accessing environment variables, or transmitting data to external endpoints. The SKILL.md explicitly recommends conservative behavior (ask user before risky actions).
Install Mechanism
No install spec and no code files — instruction-only packaging — so nothing is written to disk or fetched at install time. This is the lowest-risk install posture and is proportionate for a policy/routing skill.
Credentials
The skill requires no environment variables, credentials, or config paths. The requested scope (none) is appropriate for a policy-only router.
Persistence & Privilege
Flags show no forced always-on presence and model invocation is allowed (normal). The skill does not request elevated privileges or to modify other skills' configs. Its guidance explicitly limits autonomous risky operations.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install skill-taxonomy-router
  3. After installation, invoke the skill by name or use /skill-taxonomy-router
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.5.0
Improve query-first routing by starting from compact skill-pool summaries and reducing unnecessary deep router invocation.
v1.4.1
Remove direct references to unbundled local logging/scripts and clarify the public package as a policy-only, host-agnostic governance skill.
v1.4.0
Slim the public package to emphasize routing taxonomy and governance rules while excluding host-specific automation scripts and local state files.
v1.3.0
Add high-automation skill intake with overlap checks, inbox-only download flow, and intake state recording for safer bulk imports.
v1.2.0
Refine the download workflow to a single temporary inbox, add source/risk-signal review, and keep new-skill risk operations blocked without explicit user approval.
v1.1.0
Add usage/routing logging, backlog prioritization, overlap detection, canonical-first fusion guidance, and a hard boundary blocking new-skill risk operations without explicit user approval.
v1.0.1
Make bundled scripts portable by removing hard-coded workspace paths and keep the published package focused on the organizer/router itself.
v1.0.0
Initial release: taxonomy, risk model, routing rules, and maintenance workflow for skill selection.
Metadata
Slug skill-taxonomy-router
Version 1.5.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 8
Frequently Asked Questions

What is Skill Taxonomy Router?

Route user tasks to the most relevant skills using a layered taxonomy, risk model, and minimum-necessary-loading strategy. Use when deciding which skill to l... It is an AI Agent Skill for Claude Code / OpenClaw, with 315 downloads so far.

How do I install Skill Taxonomy Router?

Run "/install skill-taxonomy-router" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Skill Taxonomy Router free?

Yes, Skill Taxonomy Router is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Skill Taxonomy Router support?

Skill Taxonomy Router is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Skill Taxonomy Router?

It is built and maintained by Stairer (@wzk332266); the current version is v1.5.0.

💬 Comments