cost-tracker

What This Skill Does

The cost-tracker skill is a robust financial oversight tool specifically designed for OpenClaw agents interacting with the OpenRouter API. It acts as an automated auditor for your AI consumption, interfacing directly with OpenRouter's usage telemetry to capture, store, and analyze expenditure data. By utilizing a local SQLite database, it maintains a persistent historical record of every request made, the model used, and the associated cost. It enables users to move from reactive spending to proactive financial management by providing actionable insights through summary reports, budgetary guardrails, and intelligent model-swapping recommendations.

Installation

To integrate this skill into your environment, use the OpenClaw command-line interface. Ensure your OpenRouter API keys are correctly configured in your environment variables, as the tool relies on them for authenticated access to usage logs:

clawhub install openclaw/skills/skills/aiwithabidi/cost-tracker

After installation, the tool initializes a local SQLite file at {baseDir}/data/cost_tracker.db. Verify connectivity by running python3 {baseDir}/scripts/cost_tracker.py fetch to pull your initial usage data.

Use Cases

• Startup Budgeting: For developers building applications on top of LLMs, this tool prevents surprise bills by enforcing strict monthly spending caps.
• Model Optimization: Teams can identify if expensive models like GPT-4 or Claude 3.5 Opus are being used for simple tasks, allowing them to switch to more cost-effective options without sacrificing output quality.
• Trend Analysis: Agencies can correlate project milestones with specific spikes in API spending to provide transparent billing to clients.

Example Prompts

1. "What is my total spending across all models for the current week?"
2. "Set a monthly budget of $50.00 and notify me if I hit the 80% threshold."
3. "Analyze my last month of usage and suggest which models I could swap out to lower my bill while maintaining performance."

Tips & Limitations

• Data Granularity: The accuracy of the reports is dependent on how frequently you trigger the fetch command; consider setting up a cron job for automated daily updates.
• Security: Ensure your local SQLite database folder is included in your system backups but excluded from public code repositories to prevent sensitive metadata leakage.
• Limits: The tool is designed specifically for OpenRouter; it cannot track costs for direct API calls made to OpenAI, Anthropic, or Mistral outside of the OpenRouter ecosystem. Always verify your actual bill on the OpenRouter dashboard periodically to ensure consistency between internal tracking and external billing.