Metrics

What This Skill Does

The Metrics skill serves as the central intelligence engine for OpenClaw to track, normalize, and analyze performance data across any organizational or personal domain. It provides a structured architectural framework that separates raw signal ingestion from derived analytic insights. By implementing a rigorous 'Metric Contract' system, the skill ensures that every calculation is backed by clear business logic, source definitions, and owner accountability. It utilizes a modular file system structure located in ~/metrics/ to maintain scalability, allowing users to move from simple tracking to complex reporting workflows involving version-controlled formulas and automated threshold-based alerts.

Installation

To integrate this skill into your environment, execute the following command in your terminal: clawhub install openclaw/skills/skills/ivangdavila/metrics After installation, verify the setup by running the setup.md script located in the ~/metrics/ directory to initialize your memory schema.

Use Cases

• Business Growth Tracking: Monitoring North Star metrics like MRR, churn rates, or customer acquisition costs with multi-dimensional segmentation (e.g., by channel or geography).
• Product Analytics: Analyzing feature adoption rates and user workflow bottlenecks, separating event signals from behavioral insights.
• Personal Productivity: Tracking habit formation, time-on-task, or fitness metrics with custom formulas and scheduled reporting cadences.
• Operations Governance: Maintaining data quality gates and automated alerts for supply chain or budget thresholds, ensuring that any deviation triggers a pre-defined response playbook.

Example Prompts

1. "Define a new metric for 'Monthly Active Users' including a contract that specifies the numerator as unique logins and the denominator as the total user base. Store this in the registry."
2. "Update the 'Customer Lifetime Value' formula to account for the new tiered pricing structure. Please log this as version 2.0 with the reason 'Pricing Model Adjustment' and note the expected impact on historical reporting."
3. "Generate a weekly summary report for the marketing team. Ensure each section includes a 'next action' item for the channel leads and links back to the specific automation thresholds we defined for lead generation."

Tips & Limitations

• Dimension First: Always try to add a dimension (e.g., 'Region', 'Status') before creating a new independent metric to avoid data fragmentation.
• Version Control: Never perform in-place updates to formulas. Always version them in the formulas/ directory to preserve historical accuracy.
• Noise Reduction: Set up response playbooks for all automations; alerts that do not prompt a specific action should be downgraded to logs or ignored to prevent 'alert fatigue'.
• Limitation: This skill relies on the user providing consistent raw data inputs; it is not a data-scraping tool itself but a processing and reporting engine.