Compare

What This Skill Does

The Compare skill provides a sophisticated framework for OpenClaw agents to execute objective, high-stakes comparative analysis. Unlike standard summarization tools that might favor the most available data, this skill enforces 'Research Parity.' It mandates that every entity being compared is researched to an equivalent depth and quality, ensuring that a lack of information for one option does not bias the final outcome. By systematically weighting criteria, tracking research depth, and calculating confidence scores, the Compare skill transforms subjective decision-making into a rigorous, verifiable process.

Installation

To integrate this skill into your environment, run the following command within your terminal or OpenClaw interface:

clawhub install openclaw/skills/skills/ivangdavila/compare

Ensure that you have domains.md, confidence.md, traps.md, and preferences.md configured in your root directory to allow the skill to calibrate its recommendations to your specific needs.

Use Cases

This skill is ideal for scenarios where decision fatigue or information asymmetry might lead to poor choices:

• Technical Procurement: Comparing SaaS platforms or hardware where feature lists often obscure technical debt or maintenance requirements.
• Investment Due Diligence: Evaluating financial assets by normalizing research across different market reports.
• Strategic Planning: Weighing competing business initiatives based on organizational goals defined in your memory files.
• Complex Personal Decisions: Choosing between life-altering options like relocation packages or educational paths where criteria weighting is critical.

Example Prompts

1. "Compare the pricing and technical architecture of AWS Lambda vs. Google Cloud Functions for a high-concurrency event-driven pipeline."
2. "I need to choose a project management tool. Compare Trello and Jira based on my startup's preference for agile workflows and minimal setup time."
3. "Evaluate the performance differences between TypeScript and JSDoc for my current internal library, ensuring the comparison accounts for build time and type safety."

Tips & Limitations

• Weighting: Always define your criteria clearly. If you don't know what matters most, tell the agent to guide you through a prioritization phase first.
• Integrity: The skill is designed to decline requests if it cannot achieve research parity. Respect this boundary; a 'partial' answer is often more misleading than a refusal when the data is fundamentally unbalanced.
• Preferences: Use the preferences.md file to seed your long-term values. The agent will read this to automatically weight your comparison criteria, saving you from repeating your requirements in every prompt.