Model Pilot is a heuristic-driven utility that ensures you never overpay for AI tokens by matching task complexity to the most cost-effective model.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install model-pilot
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 model-pilot 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 Model Pilot?
Model Pilot is a sophisticated decision-support tool designed for the Openclaw Skills ecosystem. It prevents unnecessary expenditure on high-end LLMs by performing a quick heuristic analysis of incoming tasks. By categorizing tasks into Routine, Intermediate, or Complex tiers, it provides developers with actionable recommendations on whether a cheaper model can achieve the same quality of results.
Beyond simple classification, the skill integrates dynamic data fetching to ensure accuracy. It leverages real-time web lookups to bypass the pitfalls of hardcoded pricing, which frequently fluctuates in the AI market. This ensures that every recommendation is backed by the latest market rates and a logical cost-benefit analysis.
Model Pilot Use Cases
- When starting a potentially expensive task and needing to verify the most economical model choice.
- Instances where a user explicitly requests to save tokens or find a cheaper model alternative.
- Comparing live pricing across different AI providers for specific token counts.
- Assessing if a routine task like file formatting can be offloaded from a flagship model to a mid-tier one.
How Model Pilot Works
- The skill monitors user prompts for high-complexity triggers or explicit cost-related keywords.
- It performs a Tier Classification (Routine, Intermediate, or Complex) based on token requirements and creative needs.
- If a cheaper model is suitable, the skill triggers a live pricing fetch via web search or specific pricing URLs.
- It calculates an estimated cost using the formula: $(input_tokens / 1M \times input_price) + (output_tokens / 1M \times output_price)$.
- The agent presents a comparison to the user, highlighting potential percentage savings and quality trade-offs.
Model Pilot Setup
To integrate this utility into your workflow, install it via the Openclaw Skills manager:
clawhub install model-pilot
For the complete productivity suite by TommoT2, you can install multiple tools at once:
clawhub install model-pilot cross-check context-brief setup-doctor
Model Pilot Data Schema & Taxonomy
Model Pilot operates as a stateless recommendation engine to maintain security and speed. It organizes data according to the following taxonomy:
| Data Point | Source | Persistence |
|---|---|---|
| Task Tier | Heuristic Analysis | Session-only |
| Model Pricing | Live Web Fetch | Transient |
| Cost Estimates | Real-time Calculation | Not Persisted |
| User Preferences | Language Detection | Active Context |
It specifically avoids accessing credentials, modifying local files, or persisting user data between sessions.
Model Pilot Advanced Features
- Multi-model comparison: Simultaneously fetch and compare rates for different LLM providers.
- Dynamic Language Detection: Automatically matches the user's communication language for all cost reports.
- Quality Trade-off Warnings: Provides honest assessments if a cheaper model recommendation might result in lower output quality.
- Token-Specific Heuristics: Automatically flags tasks requiring <1000 output tokens with low creativity as Tier 1 (Routine) for immediate cost reduction.
SKILL.md
Loading
Related Openclaw Skills
GitLab API Skill
A powerful Node.js client providing a comprehensive interface for interacting with the GitLab REST API v4.
Google Trends Keyword Trend Analysis
A powerful AI tool for analyzing Google search interest trends and regional keyword popularity worldwide.
King Wen Hexagrams
An AI-driven divination system based on King Wen’s 64 hexagrams for trend analysis, risk assessment, and daily guidance.
Google Trends Real-Time Analysis
An AI-powered skill to query and analyze real-time Google Trends data for discovering breakout search topics across global markets.
engramai
A neuroscience-grounded memory system for AI agents that implements cognitive learning patterns like ACT-R activation and Hebbian learning.
Xiaohongshu Editor Pick Style
A professional writing skill for Xiaohongshu designed to produce clean, structured, and editorial-grade content for high-value social media posts.
Featured*
