lenny-skillpack-creator

To convert unstructured skill descriptions into an agent-executable skill pack, you refactor the raw content into a standardized format defining clear scope, input/output contracts, workflows, and quality gates. This process creates self-contained directories including SKILL.md, references, and optional scripts for AI agents.

An agent-executable skill pack is a standardized directory format containing SKILL.md, references, and scripts that enables AI agents like Claude Code or Codex to execute methodologies step-by-step. You need it when transforming insight-heavy documentation into callable, reusable AI agent capabilities.

Yes, you can use existing Refound or Lenny skills as input. The skill conversion process specifically takes these insight-heavy, semi-structured skill descriptions and refactors them into structured, agent-executable formats with clear inputs, outputs, and workflows.

The best way to structure workflows and quality gates for AI agent skills is by standardizing them within a SKILL.md file during the agent pack generation process. This defines the execution scope, input/output contracts, and quality checkpoints required for reliable step-by-step methodology execution.

No, the generated agent skill pack supports English output only for all generated artifacts. The standardized skill directories, including the SKILL.md file, references, and optional scripts, are produced with English content to ensure consistent agent execution.

Input/output contracts are important when creating agent skills because they define the exact scope and data boundaries for the AI agent. Establishing these contracts during skill conversion ensures the methodology is executed reliably step-by-step without ambiguity in the consumed inputs or produced outputs.