LLM Switcher for Openclaw

A robust tool to change OpenClaw's global default AI model with built-in availability checks and automated configuration management.

winlinvip
v1.0.0
Mar 3, 2026
0
496
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install llm-switcher

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install llm-switcher 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 LLM Switcher?

The LLM Switcher skill is designed to manage the primary intelligence source for your agentic environment. It automates the complex task of modifying global configuration files while ensuring that any new model selection is fully functional before the change is finalized. By integrating with the broader ecosystem of Openclaw Skills, this tool minimizes downtime and prevents configuration errors that could lead to broken workflows.

This skill provides a guided, interactive experience that moves from discovery to validation, and finally to deployment. It handles the nuances of model aliases and provider IDs, making it simple to transition between different LLMs for different project requirements without manual file editing.

LLM Switcher Use Cases

  • Transitioning the global default model from a lightweight testing model to a high-reasoning production model.
  • Validating newly added model aliases to ensure they are correctly configured and responsive.
  • Safely updating environment-wide defaults across multiple Openclaw Skills and agents.
  • Troubleshooting model connectivity issues by performing session-level overrides before applying permanent changes.

How LLM Switcher Works

  1. The skill retrieves and displays the current primary model and all configured model aliases from the system configuration.
  2. The user is prompted to choose a new model from the validated list of available options.
  3. The skill performs a session-only override to test the selected model's responsiveness with a live status query.
  4. If the live test passes, the skill updates the global primary model configuration using internal CLI tools.
  5. Upon verification of the successful config write, the skill requests permission to restart the gateway.
  6. The gateway is restarted, and the final system status is verified to ensure the new model is active.

LLM Switcher Setup

To use this skill, ensure your environment has at least one model configured within the Openclaw configuration. You can verify your current setup using the following commands:

openclaw config get agents.defaults.models
openclaw config get agents.defaults.model.primary

No additional installation is required if the base gateway and CLI tools are already active.

LLM Switcher Data Schema & Taxonomy

This skill interacts with the core configuration schema to manage model persistence:

Configuration Path Purpose
agents.defaults.models Source of truth for all available model aliases and IDs.
agents.defaults.model.primary The global target updated by this skill.
session_status Used as a temporary metadata bucket for model validation.

All changes are verified against the local configuration state before being reported as successful.

LLM Switcher Advanced Features

  • Session-level availability checks that prevent breaking the global configuration with an invalid model.
  • Interactive alias resolution that supports both short-form names and full provider identifiers.
  • Automated gateway lifecycle management, including status verification and error recovery suggestions.
  • Direct integration with Openclaw Skills workflows to provide a consistent management interface.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*