Skill creator

Skill nainishshafi/developer-productivity-skills/.github/skills/skill-creator

Reusable AI skills for Claude Code and GitHub Copilot — scan READMEs, sync forks, create skills, lint Python, and scan for security vulnerabilities

Install

npx -y skills add nainishshafi/developer-productivity-skills --skill skill-creator

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Use when the user asks to "create a new skill", "add a skill", "write a skill", "scaffold a skill", "make a claude skill", "build a skill", or wants to create a reusable Claude Code skill from scratch. Also use when the user describes a repeatable workflow they want to package as a skill.

SKILL.md

3.5 KB, as published. Nobody here has run it

Skill Creator

Scaffold and write new Claude Code skills following established conventions. Handles both reference/knowledge skills (rich documentation that Claude loads as context) and task/action skills (step-by-step workflows with scripts).

When to Use This Skill

User wants to package a repeatable workflow as a skill
User asks to "create", "add", "write", or "scaffold" a skill
User describes something they want Claude to do consistently across sessions
User wants to share a skill across projects or with others

Workflow

Step 1 — Classify the Skill Type

Ask the user:

What should the skill do? (one sentence)
Does it perform an action or provide knowledge?
- Action: runs scripts, executes steps, interacts with tools → Task skill
- Knowledge: patterns, templates, conventions, reference material → Reference skill

Decision:

Reference skill → single rich SKILL.md only (no scripts required)
Task skill → SKILL.md + references/ + scripts/

Step 2 — Gather Details

Collect from the user:

Skill name — kebab-case, lowercase (e.g., lint-python, deploy-staging)
Trigger phrases — what will the user say to invoke it? (list 5–8 examples)
Target location — .github/skills/ (this project) or ~/.claude/skills/ (all projects)
For reference skills: topic areas, frameworks, code patterns to cover
For task skills: step-by-step actions, prerequisites, scripts needed (bash or python?)

Step 3 — Scaffold the Folder

Run the scaffold script to create the directory structure and boilerplate:

[ -d .venv ] || python -m venv .venv
PYTHON=$(if [ -f .venv/Scripts/python ]; then echo .venv/Scripts/python; else echo .venv/bin/python; fi)
$PYTHON .github/skills/skill-creator/scripts/scaffold-skill.py <skill-name> [base-path]

Default base-path is .github/skills/
Script outputs created file paths to stdout (one per line)
If the skill folder already exists, the script exits with a warning — do not overwrite

Step 4 — Populate the Files

Use references/skill-anatomy.md as your writing guide.

For a reference skill, write a rich SKILL.md with these sections:

When to Use This Skill — bullet list of scenarios
Core Concepts — fundamental ideas with explanations
Implementation Patterns — named patterns with code examples (3–10+ patterns)
Best Practices — do's and don'ts
Common Pitfalls — what to avoid and why
Resources — links and further reading

For a task skill, write:

SKILL.md — frontmatter + step-by-step workflow (reference references/ and scripts/ by path)
references/<skill-name>-reference.md — supporting documentation, patterns, and specs
scripts/<script>.py or scripts/<script>.sh — executable logic with error handling

Confirm all created files with the user before finishing.

Additional Resources

references/skill-anatomy.md — Complete frontmatter field reference, skill type templates, description writing guide, script conventions, naming rules
scripts/scaffold-skill.py — Creates folder structure and boilerplate files for a new skill