openclaw-cost-tracker

What This Skill Does

The openclaw-cost-tracker is a specialized analytical utility designed for OpenClaw users who require granular visibility into their AI expenditure. By directly querying session logs, the skill identifies the specific costs associated with every interaction, categorized by model type and date. It automates the extraction and calculation of these costs, rendering them into readable summaries. This allows power users to manage their budgets effectively, optimize model selection based on cost-efficiency, and identify potential spikes in usage before they impact financial plans. The skill is particularly useful for teams or individuals managing multiple AI workflows across different models, such as Claude-opus or GPT-4o, where costs can fluctuate based on token usage.

Installation

To integrate this tool into your OpenClaw environment, ensure the jq utility is installed on your host system (via 'brew install jq' or 'apt install jq'). Once jq is ready, execute the installation command: 'clawhub install openclaw/skills/skills/vincentqiu/cost-report'. The skill will automatically point to your local configuration directory at '~/.openclaw/agents/'. No complex API keys are required, as the tool relies on your existing, locally stored session logs.

Use Cases

• Financial Oversight: Regular monitoring of daily and weekly spending to adhere to strict AI research budgets.
• Model Optimization: Comparing the cost-per-result for different LLMs to determine which model provides the best value for specific tasks.
• Reporting for Stakeholders: Generating professional cost summaries that can be copied directly into Discord or Slack channels to inform project leads of infrastructure consumption.
• Trend Analysis: Identifying periods of high activity and forecasting future spending based on month-over-month growth data.

Example Prompts

1. "OpenClaw, run the cost tracker for this week and show me how much I've spent compared to last week."
2. "Can you give me a breakdown of my costs for the month of January 2026, categorized by the model I used?"
3. "Summarize yesterday's usage costs and post the result in the standard Discord format."

Tips & Limitations

• Data Integrity: The tracker relies on the contents of the 'message.usage.cost.total' field in your logs. If your OpenClaw configuration does not explicitly log these fields, the output may be incomplete.
• Performance: Because this script iterates through your session history, very large log directories might cause a slight delay during execution.
• Local Processing: Since all analysis is done locally on your machine, your private usage data is never transmitted to an external server or third-party API, ensuring maximum privacy.