A technical advisor for OpenClaw agents that identifies token waste and provides mathematical cost-saving recommendations.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install optimoclaw-skill
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 optimoclaw-skill 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 OptimoClaw Token Optimizer?
OptimoClaw Token Optimizer is a sophisticated diagnostic utility designed to help developers control the operational costs of their AI agents. By interfacing directly with the OpenClaw CLI, this skill analyzes live configuration files and session usage metrics to pinpoint exactly where tokens are being wasted. It serves as an essential bridge between agent performance and budget management, ensuring that users can maintain high-quality outputs while minimizing unnecessary API spend.
As one of the most specialized Openclaw Skills for financial management, it provides a transparent view of model expenses across providers like Anthropic and OpenAI. The skill doesn't just suggest changes; it provides the exact math and logic behind every recommendation, allowing developers to make informed decisions about model routing, heartbeat frequencies, and context window utilization.
OptimoClaw Token Optimizer Use Cases
- Reducing monthly API bills by optimizing heartbeat models and frequency.
- Improving cache efficiency to take advantage of provider discounts on repeated inputs.
- Tuning context compaction thresholds to prevent agents from becoming sluggish or expensive over long sessions.
- Benchmarking different model tiers to find the most cost-effective routing for routine tasks.
How OptimoClaw Token Optimizer Works
- The skill initiates data collection by executing
openclaw config getto map current agent settings and model assignments. - It runs
openclaw status --usage --jsonto retrieve real-time telemetry on token counts, cache ratios, and context utilization. - It processes these metrics against a built-in rate card for various LLM providers to calculate current spend.
- It evaluates the data against 10 specific optimization levers, such as memory file scope and subagent concurrency.
- The skill generates a comprehensive report detailing identified waste, recommended values, and the specific shell commands to apply the updates.
OptimoClaw Token Optimizer Setup
To utilize this skill, ensure the OpenClaw CLI is installed and accessible in your environment. No separate installation is required beyond providing the agent with the skill definition.
# Verify OpenClaw CLI accessibility
openclaw --version
# The agent will automatically use the following commands to gather data:
openclaw config get 'agents' --json
openclaw status --usage --json
OptimoClaw Token Optimizer Data Schema & Taxonomy
The skill organizes and analyzes data based on the standard OpenClaw configuration hierarchy and session metadata:
| Data Point | Description |
|---|---|
| Agent Defaults | Primary models, heartbeat settings, and search batch limits. |
| Usage Telemetry | Cache reads vs. writes and context window percentages. |
| Rate Card | A reference table of costs per million tokens for input, output, and cache. |
| Optimization Profiles | Pre-defined configurations for Lean, Balanced, and Quality performance modes. |
OptimoClaw Token Optimizer Advanced Features
- Profile-based presets that allow users to toggle between aggressive cost-saving and high-fidelity quality modes.
- Local model routing recommendations for Ollama to achieve zero-cost heartbeats and summaries.
- Cache efficiency auditing to identify sessions where context loading settings are causing redundant writes.
- Per-agent override analysis to distinguish between global defaults and specific agent requirements.
SKILL.md
Loading
Related Openclaw Skills
Meeting Extractor
An AI-powered tool that converts messy meeting transcripts into structured summaries, action items, and key decisions.
Network Diag
A versatile command-line utility for verifying host connectivity, port status, and DNS resolution.
Adaptive Brain
Adaptive Brain is a self-evolving agent system that functions as a behavioral immune system, detecting error patterns and automatically optimizing agent performance over time.
Task Board Skill
A lightweight, CLI-driven kanban board for managing tasks and project workflows directly from your terminal.
Agent Plan Manager
A structured framework for managing AI agent execution workflows through Markdown-based planning and tracking.
Amazon EKS Security Hardening Tool
An automated security configuration generator that hardens Amazon EKS clusters based on industry-standard CIS Benchmarks.
Featured*
