smart-model-switching-glm
Auto-route tasks to the cheapest z.ai (GLM) model that works correctly. Three-tier progression: Flash → Standard → Plus/32B. Classify before responding. FLASH (default): factual Q&A, greetings, reminders, status checks, lookups, simple file ops, heartbeats, casual chat, 1–2 sentence tasks, cron jobs. ESCALATE TO STANDARD: code >10 lines, analysis, comparisons, planning, reports, multi-step reasoning, tables, long writing >3 paragraphs, summarization, research synthesis, most user conversations. ESCALATE TO PLUS/32B: architecture decisions, complex debugging, multi-file refactoring, strategic planning, nuanced judgment, deep research, critical production decisions. Rule: If a human needs >30 seconds of focused thinking, escalate. If Standard struggles with complexity, go to Plus/32B. Save major API costs by starting cheap and escalating only when needed.
Why use this skill?
Optimize OpenClaw agent costs with smart model routing. Automatically switch between GLM Flash, Standard, and Plus tiers based on task complexity.
Install via CLI (Recommended)
clawhub install openclaw/skills/skills/princnl/smart-model-routing-for-zaiWhat This Skill Does
The smart-model-switching-glm skill is a sophisticated routing engine designed to optimize API usage and model performance within the OpenClaw ecosystem. By implementing a three-tier progression logic (Flash → Standard → Plus/32B), the skill ensures that every request is processed by the most cost-effective model capable of handling the task. It performs automatic classification of user intent, routing simple queries like heartbeats or casual chat to Flash, general development and analysis to Standard, and complex architectural decisions to Plus/32B.
Installation
To integrate this routing engine, run the following command in your terminal:
clawhub install openclaw/skills/skills/princnl/smart-model-routing-for-zai
Once installed, the agent will automatically intercept relevant tasks, provided the skill is enabled in your configuration.
Use Cases
- Automated Monitoring: Use for lightweight cron jobs and status updates to minimize infrastructure costs.
- Dynamic Development: Seamlessly switch between fast code generation and deep architectural refactoring when the complexity of a PR increases.
- Research Synthesis: Automatically escalate from simple lookup to deep, multi-source research tasks.
- Cost-Effective Chat: Keep daily conversations and greetings on efficient tiers, escalating to heavier reasoning only when the user requires deeper analysis or complex table generation.
Example Prompts
- "Check the current disk usage on the production server and report the status."
- "Write a Python script to scrape this website and store the results in a structured JSON format."
- "Evaluate the trade-offs between a monolithic and microservices architecture for our new high-concurrency payment system."
Tips & Limitations
- The 30-Second Rule: As a heuristic, if you feel the task would take a human longer than 30 seconds of concentrated effort, don't force it to Flash. Let the skill route it to Standard.
- Complexity Scaling: If a task on the Standard tier is consistently failing or producing shallow results, manually force the sub-agent call to the Plus/32B model.
- Avoid Over-Escalation: Do not default to Plus/32B for simple factual lookups; this drastically increases API latency and unnecessary costs.
- Context Awareness: Ensure the input task is descriptive, as the classifier relies on the prompt content to make its routing decision.
Metadata
Not sure this is the right skill?
Describe what you want to build — we'll match you to the best skill from 16,000+ options.
Find the right skillPaste this into your clawhub.json to enable this plugin.
{
"plugins": {
"official-princnl-smart-model-routing-for-zai": {
"enabled": true,
"auto_update": true
}
}
}Tags(AI)
Flags: external-api