Creates reusable AI skills with SKILL.md instructions and tested drafts.
npx skills add https://github.com/apappas57/sinew --skill skill-creator-apappas57
Please help me install this Agent Skill. Skill: skill-creator Source: https://github.com/apappas57/sinew/tree/main/skills/core/skill-creator Command: npx skills add https://github.com/apappas57/sinew --skill skill-creator-apappas57
Skill Creator helps you turn an idea, rough workflow, or existing skill draft into a reusable skill that Claude can trigger reliably and improve over time.
Use the skill creator to define the workflow you want, generate a draft skill, test it with realistic prompts, and refine it until the results are reliable.
High-intent search queries and answers about installing and using this skill.
To write SKILL.md instructions, structure your workflow into clear triggering guidance, step-by-step instructions, and optional supporting resources. Skill Creator drafts these components from your initial idea and refines them through evaluation feedback.
To create reusable AI skills, draft your initial workflow idea into structured instructions with clear trigger conditions, then test against realistic prompts to refine activation reliability. Skill Creator handles this end-to-end process from ideation to a tested final draft.
Iterative skill improvement involves testing outputs against realistic prompts, comparing results, reviewing feedback, and revising instructions until performance is consistent. Skill Creator benchmarks outputs and applies feedback to refine trigger conditions and instructions.
Benchmarking AI skills requires running evaluation prompts, inspecting quantitative grading results, and optimizing the skill description for accurate real-task activation. Skill Creator provides built-in evaluation prompts and grading workflows to measure and tune this reliability.
Trigger optimization for AI skills involves tuning the skill description so Claude activates it accurately during real tasks. Skill Creator supports description tuning by benchmarking activation rates against test prompts and iterating on the results.
No external frameworks are required to start authoring AI skills. Skill Creator operates independently with no dependencies, allowing you to define workflows, generate draft instructions, and build evaluation prompts directly.
dlt-hubScaffold trigger evaluation setups for toolkit skills with test workspaces and queries.
JansenAnalyticsMonitor application logs and stderr to detect errors, crashes, and exceptions.
duanbiao2000Evaluate LLM applications with automated metrics, human judgments, and benchmarking.
jaimedhenriquesDraft skill instructions, trigger conditions, and evaluation metadata.
