Skills skill-creator

📦

skill-creator

Name: skill-creator
Author: ZhanlinCui

Safe

Create custom skills for Claude

Also available from: 92Bilal26,AbdullahMalik17,langchain-ai,ChinchillaEnterprises,DaoThiHuong2111,vercel-labs,davila7,BarisSozen,ArtemisAI,Joseph OBrien,BOM-98,7Spade,AstraBit-CPT,Doyajin174,CesarAugustusGroB,BenderFendor,Cam10001110101,Joseph OBrien,7Spade,openai,ComposioHQ,92Bilal26,Azeem-2,DavidDworetzky

Building effective AI skills requires understanding context management, progressive disclosure, and reusable resource design. This skill provides a complete framework for creating modular skills that extend Claude's capabilities without wasting context window space.

Supports: Claude Codex Code(CC)

🥈 80 Silver

Download the skill ZIP

Upload in Claude

Go to Settings → Capabilities → Skills → Upload skill

Toggle on and start using

Test it

Using "skill-creator". I want to create a PDF editing skill. What should I include?

Expected outcome:

Based on concrete use cases like 'rotate this PDF' or 'extract text from page 3':
1. scripts/rotate_pdf.py - deterministic PDF rotation utility
2. scripts/extract_text.py - text extraction with page range support
3. references/pdf-api.md - reference for PDF library methods
4. SKILL.md body with task-based structure: Quick Start, Rotate PDFs, Extract Text, Merge PDFs
Keep SKILL.md under 500 lines and link to reference files for detailed API documentation.

Using "skill-creator". How do I write effective frontmatter for my skill?

Expected outcome:

Frontmatter example for a docx skill:
---
name: docx-editor
description: Comprehensive document creation, editing, and analysis with support for tracked changes, comments, formatting preservation, and text extraction. Use when Claude needs to work with professional documents (.docx files) for: (1) Creating new documents, (2) Modifying or editing content, (3) Working with tracked changes, (4) Adding comments, or any other document tasks
---
Key: Include both what the skill does AND specific triggers for when to use it.

Security Audit

Safe

v1 • 2/24/2026

All 136 static findings were evaluated and dismissed as false positives. The 'external_commands' detections are markdown code fence backticks in documentation, not shell execution. The 'weak cryptographic algorithm' findings reference documentation examples, not actual crypto code. The 'filesystem' detections are legitimate zipfile usage for packaging skills into distributable .skill files. This is a documentation skill with utility scripts for skill initialization and packaging.

Files scanned

978

Lines analyzed

findings

Total audits

No security issues found

Audited by: claude

Quality Score

Architecture

100

Maintainability

Content

Community

100

Security

Spec Compliance

What You Can Build

Developer building a code review skill

A developer wants to create a skill that helps Claude conduct thorough code reviews following their team's standards. They use skill-creator to structure the SKILL.md with review workflows, add reference docs for coding standards, and package it for team distribution.

Data analyst creating a BigQuery skill

An analyst needs Claude to query their company's BigQuery tables efficiently. Using skill-creator, they organize schema documentation in references/, design task-based workflows for common queries, and keep SKILL.md lean with progressive disclosure.

Brand team building a style guide skill

A brand manager wants Claude to create on-brand content. They use skill-creator to structure brand guidelines, store logo and template files in assets/, and package the skill so all team members get consistent brand-compliant outputs.

Try These Prompts

Beginner: Create a simple documentation skill

I want to create a skill that provides guidelines for writing technical documentation at my company. Help me use the skill-creator workflow to set up the directory structure, write the SKILL.md frontmatter and body, and organize our documentation standards in a references file.

Intermediate: Build a workflow-based skill

I need to create a skill for processing customer support tickets that involves multiple steps: categorization, response drafting, and escalation routing. Using skill-creator, help me design a sequential workflow, decide what belongs in scripts versus references, and implement the SKILL.md body with clear decision points.

Advanced: Create a skill with bundled resources

Build a comprehensive data analysis skill that includes: Python scripts for common data transformations in scripts/, database schema documentation in references/, and Jupyter notebook templates in assets/. Guide me through analyzing concrete use cases, planning reusable resources, and packaging the final skill.

Expert: Refactor an existing skill using progressive disclosure

I have an existing skill that's too large and loads too much context. Using skill-creator principles, help me refactor it to use progressive disclosure: identify what should stay in SKILL.md versus move to reference files, restructure for domain-specific organization, and optimize for context window efficiency.

Best Practices

Keep SKILL.md body under 500 lines and use progressive disclosure to move detailed content into reference files
Write frontmatter descriptions that include both capabilities and specific trigger scenarios for when Claude should use the skill
Use imperative verb forms in instructions and organize by workflow or task depending on your skill's purpose

Avoid

Including README.md, CHANGELOG.md, or other auxiliary documentation that clutters the skill without helping Claude execute tasks
Putting all information in SKILL.md instead of splitting large content into reference files loaded only when needed
Writing vague frontmatter descriptions that only explain what the skill does without specifying when to trigger it

Frequently Asked Questions

What is the maximum size for a skill?

There is no hard limit, but keep SKILL.md under 500 lines (~5000 words) to avoid context bloat. Use progressive disclosure to move detailed documentation into reference files that Claude loads only when needed. The total skill package size is flexible since scripts can execute without loading into context.

When should I include scripts versus writing instructions?

Include scripts when the same code is rewritten repeatedly or when deterministic reliability is critical. Write instructions when Claude needs to adapt to context or when multiple valid approaches exist. Scripts are token-efficient and can execute without context loading, but may still need to be read for patching.

How do I decide between workflow-based and task-based structure?

Use workflow-based structure for sequential processes with clear step-by-step procedures (like PDF filling). Use task-based structure for tool collections with different operations (like a design toolkit). Most skills combine patterns: start with task categories, then add workflow guidance for complex operations within each task.

Can skills include external API integrations?

Yes. Include API documentation in references/, authentication setup scripts in scripts/, and clear usage instructions in SKILL.md. Document rate limits, error handling, and provide example requests. Ensure the skill clearly states when network access is required.

What is a .skill file and how is it different from a skill folder?

A .skill file is a zip archive with a .skill extension containing all skill files. It is the distributable format for sharing skills. The package_skill.py script validates your skill first, then creates the .skill file maintaining the directory structure. Users receive the .skill file and extract it to use the skill.

How do I iterate on a skill after users provide feedback?

Follow the iteration workflow: (1) Use the skill on real tasks, (2) Notice where Claude struggles or is inefficient, (3) Identify specific updates needed in SKILL.md or resources, (4) Implement changes and test again. Users often provide feedback immediately after use while context is fresh.