Create and evaluate Claude Skills using SKILL.md and Python scripts.
npx skills add https://github.com/onzhang/BookNexus --skill skill-creator-onzhang
Please help me install this Agent Skill. Skill: skill-creator Source: https://github.com/onzhang/BookNexus/tree/main/skills/skill-creator Command: npx skills add https://github.com/onzhang/BookNexus --skill skill-creator-onzhang
Create new Claude Skills, modify and improve existing ones, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize a skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.
Provide a new skill outline in SKILL.md, then run the evaluation loop to draft, test, and improve the skill until it meets your criteria.
High-intent search queries and answers about installing and using this skill.
Automating Claude skill creation involves drafting a SKILL.md file with frontmatter, then running integrated Python scripts like run_eval to test, benchmark, and refine the skill iteratively until it meets your criteria.
Optimizing skill descriptions for triggering accuracy requires iterating on the SKILL.md frontmatter using the improve_description script, which benchmarks performance against real-world prompts to boost activation reliability.
Running eval loops to benchmark Claude skill performance uses the repository's evaluator tooling to execute train and test cycles, measuring outcomes with variance analysis to drive iterative improvements.
Creating and refining Claude skills requires Python and the pyyaml dependency to execute the repository's automation scripts for drafting, testing, and optimizing skill workflows.
Benchmarking skill performance with variance analysis is supported by the integrated evaluator tooling, which runs end-to-end evaluation loops to measure outcomes and identify areas for skill refinement.
SKILL.md frontmatter must include the skill name and description to guide activation, while the body captures intent and provides robust instructions that ensure accurate triggering in real-world prompts.
tuliosouza99Create and iteratively improve SKILL.md-based agent skills with Python eval scripts.
DayMoonDevelopmentAutomate creation, evaluation, and iteration of Claude skills with SKILL.md.
Patasse97Validate Claude Skills structure, Python scripts, and quality scores in CI/CD workflows.
wtgoku-createIteratively refine SKILL.md descriptions through automated trigger evaluations and benchmark outputs.
