Create, refine, and benchmark AI skills with structured workflows.
npx skills add https://github.com/ZSHYC/ASVSim_zsh111111 --skill skill-creator-zshyc
Please help me install this Agent Skill. Skill: skill-creator Source: https://github.com/ZSHYC/ASVSim_zsh111111/tree/main/.claude/skills/skill-creator Command: npx skills add https://github.com/ZSHYC/ASVSim_zsh111111 --skill skill-creator-zshyc
This Skill streamlines the entire process of creating, improving, and testing AI skills, making it easier for anyone to build powerful custom AI capabilities.
SKILL.md, create test cases, run evaluations, and optimize the description so Claude Code reliably uses it when needed.Use the skill-creator to help me build a new skill that summarizes meeting notes.
High-intent search queries and answers about installing and using this skill.
To build and test custom AI skills, use structured workflows for intent capture, prompt engineering, and test case generation, enabling iterative refinement through performance benchmarking and execution analysis.
Optimizing AI skill descriptions involves automated description tuning based on evaluation metrics, ensuring the AI reliably triggers and uses the correct skill when needed during task execution.
Yes, you can generate test cases and run evaluations by integrating with evaluation frameworks, executing deterministic and qualitative assessments to measure skill performance and identify areas for iterative improvement.
No specific dependencies are required to start with AI skill development. The process supports end-to-end creation using included scripts, references, and assets to facilitate intent capture and code generation.
Custom AI skills fail to trigger reliably due to unoptimized descriptions. Running performance benchmarking and automated description tuning refines the intent matching and triggering accuracy.
yu13130122297Identify uncertainties in research framing, datasets, metrics, and baselines.
davduncIteratively modify targets and select improvements using numeric metrics or LLM-as-judge evals.
locnbk2002Create, refine, and evaluate AI skills with SKILL.md authoring and test execution.
EpiphyticCreate isolated git worktrees and structured prompts for parallel agent tasks.