Create, test, and iterate Claude skills with SKILL.md structure.
npx skills add https://github.com/rolopez23/agent-harness --skill skill-creator-rolopez23
Please help me install this Agent Skill. Skill: skill-creator Source: https://github.com/rolopez23/agent-harness/tree/main/.claude/skills/skill-creator Command: npx skills add https://github.com/rolopez23/agent-harness --skill skill-creator-rolopez23
This Skill-creator workflow provides a structured, repeatable process to design, test, and improve Claude skills, including a YAML frontmatter entry point, a rich Markdown body, and optional resources.
Start by drafting a new skill, fill in SKILL.md with a compelling name and description, and run the built-in eval/iterate loop to refine its triggers and performance.
High-intent search queries and answers about installing and using this skill.
To create and iterate on Claude skills, use a structured workflow that guides drafting frontmatter, writing body instructions, and running an evaluation loop to refine trigger accuracy and performance.
The best way to benchmark prompt engineering skills is to run a guided evaluation loop that drafts test prompts, reads results, and iterates on instructions to ensure trigger accuracy and safety.
Yes, the skill creation workflow supports packaging optional bundled resources including scripts, references, and assets to support deterministic tasks, rich context, and output templates.
Yes, you can use YAML frontmatter to define mandatory name and description fields that guide activation and usage of automation workflows within a comprehensive Markdown body.
Inaccurate triggering during evaluation occurs when frontmatter descriptions and body instructions lack specificity, requiring iteration through the guided evaluation loop to refine performance.
Fudan-SMI-labPreinstall PyYAML before torch-npu to prevent transitive dependency failures.
ljk486Design reusable agent skills with SKILL.md structure and YAML frontmatter.
pipixia-labsGuide creation of SKILL.md files and coordinate evals, benchmarks, and iteration.
charly-vibesExtract reusable prompt templates from AI conversations by analyzing successful patterns.
