oa-agent-library-builder

To create reusable AI agent templates, this Skill analyzes completed agent workspaces to identify repeatable tasks and extracts key parameters into a structured JSON template. It evaluates task categories, necessary tools, and complexity to generate a template for agent library integration.

Pattern analysis for AI agent templates works by evaluating completed agent runs to determine task category, required tools, complexity, and atomicity. It then uses these patterns to automate template generation and assigns the optimal Claude model for cost efficiency.

Yes, you can automate Claude model selection during agent template generation. The Skill assigns the most cost-effective Claude model—Haiku, Sonnet, or Opus—based on the extracted complexity and atomicity of the identified task pattern.

The best way to save a successful agent run is to use a template extraction process that identifies repeatable tasks and generates a structured JSON template file. This file adheres to a strict schema for seamless agent library integration.

Yes, a strict JSON schema is required for agent library integration. The Skill generates a structured JSON template file that captures the extracted task patterns, tools, and model hints to ensure the new agent is reusable and standardized.

You should avoid automated agent template extraction when a completed agent run involves highly unique, non-repeatable tasks. The process relies on identifying consistent patterns, so one-off complex workflows without reusable components are not suitable for JSON template generation.