- Home
- Skills
- Nilecui
- Skillsbase
- Skill Writer
skill-writer_skill
- Python
20
GitHub Stars
1
Bundled Files
2 months ago
Catalog Refreshed
4 months ago
First Indexed
Readme & install
Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.
Installation
Preview and clipboard use veilstrat where the catalogue uses aiagentskills.
npx veilstrat add skill nilecui/skillsbase --skill skill-writer- SKILL.md9.4 KB
Overview
This skill guides you through creating well-structured Agent Skills for Claude Code, focusing on scope, location, frontmatter, content structure, and validation. It helps you design, write, test, and debug Skills so they reliably activate and behave as expected. Use it when you need hands-on guidance to author or improve a Skill.
How this skill works
I ask clarifying questions to define a single capability and recommend a precise name and location (personal vs project). I generate the required frontmatter, clear instructions, examples, and optional supporting files, then provide a validation checklist and debugging steps. Finally, I give testing steps and suggestions to refine triggers so the Skill activates correctly.
When to use it
- Creating a new Agent Skill from scratch
- Converting prompts or workflows into a reusable Skill
- Improving discovery by refining the frontmatter and description
- Troubleshooting a Skill that doesn’t activate or behaves incorrectly
- Designing project-shared Skills versus personal experimental ones
Best practices
- One Skill, one focused capability — avoid mega-Skills
- Use a concise, trigger-rich description that states what and when to use it
- Name the Skill with lowercase letters, numbers, and hyphens only
- Provide step-by-step instructions and concrete examples, not vague guidance
- Put advanced details in separate reference or examples files for progressive disclosure
Example use cases
- Author a PDF extraction Skill that triggers on phrases like “extract text from PDF” and lists .pdf in the description
- Create a project-level data-processor Skill with helper scripts and usage examples
- Convert a manual workflow into a Skill that runs a validation script and returns results
- Refine a code-review Skill to be read-only by restricting allowed tools
- Troubleshoot activation by iterating on trigger words and testing with relevant queries
FAQ
Use personal for experimental, user-specific workflows and project for shared, versioned Skills that follow team conventions.
What makes a good description?
Combine what the Skill does, when to use it, and common trigger phrases or file types so the agent reliably matches user queries.