find-skills

To find agent skills for specific tasks, you can search the skills registry using natural language queries like "how do I do X". The discovery process translates your request into targeted searches and surfaces relevant community-tested candidates across domains like web development, testing, and design.

The best way to verify agent skill quality is to check install counts, source reputation, and repository signals. The discovery process advises evaluating these credibility metrics before recommending a skill, ensuring you adopt reliable and community-tested solutions.

To install a recommended agent skill, use the explicit install commands provided during the discovery process. It offers clear installation guidance, including optional global installation commands and direct source links, allowing you to quickly extend agent functionality.

Yes, you can search for existing skills instead of reimplementing capabilities. The discovery tool helps you find installable agent skills to save time and surface community-tested solutions, preventing unnecessary manual development for common tasks like React performance optimization.

Agent skill discovery works across multiple technical domains such as web development, testing, design, and deployment. It applies to user requests for extending agent functionality by searching the central registry and surfacing matching candidates regardless of the specific domain.

A limitation of using a skills registry is reliance on external source reputation and install counts as quality indicators. You must manually verify repository signals and evaluate the credibility of candidate skills before proceeding with the provided installation commands.