Dev & Engineering

Skill Creator

A guided skill-authoring workflow that helps you build well-structured, context-efficient custom Skills for AI agents, with init and packaging scripts included.

70/ 100
Strongly recommended

As well evidenced as a documentation review can get, with no uncontrolled major risk; still check your environment and data sensitivity.

See how it was scored ↓
Works as-is in
Codex · Claude Code
Stars
★ 2.2k
Last updated
12d ago
License
MIT
skill-authoringskill-creatorprogressive-disclosureprompt-engineering
+2agent-workflowsskill-packaging

What does this skill do, and when should you use it?

skill-creator is a meta-skill bundled with Open Cowork that guides users—or the AI itself—through creating new Skills: modular packages that extend an agent with specialized knowledge, workflows, and tools. It teaches the standard skill anatomy (SKILL.md plus optional scripts/references/assets directories), core principles like conciseness and appropriate degrees of freedom, and progressive disclosure for context management. The process runs from understanding concrete usage examples through initialization, editing, and validation-based packaging into a distributable .skill file. It performs no business tasks itself; its job is turning your domain knowledge into a reusable skill.

  • Interviews you with concrete examples to define the skill's functionality and trigger scenarios
  • Analyzes examples to plan reusable scripts/, references/, and assets/ resources
  • Runs scripts/init_skill.py to scaffold a skill with frontmatter template and example directories
  • Guides writing of SKILL.md and resources using conciseness and freedom-level guidelines
  • Runs scripts/package_skill.py to validate and package the skill into a .skill file
  • Supports an iterate loop based on real usage after testing
Good fit
  • A developer who wants to encode internal database schemas or API docs into a team-shared query skill
  • An operations person who repeatedly asks for slide decks and wants templates stored as assets in a dedicated skill
  • An engineer with repetitive scripting tasks (e.g., PDF rotation) who wants scripts/ files instead of rewriting code each time
  • An Open Cowork user building custom office-automation skills for their sandboxed workspace
Not a fit
  • Users without any Agent Skills-compatible client—the artifacts and process are built around the SKILL.md ecosystem
  • Casual users who just want one-off task completion and don't need reusable, packaged workflows

How do you install this skill?

Before you start
Your agent needs
  • Shell / CLI
  • Local filesystem
Install first
  • Python (for init_skill.py / package_skill.py)

Pre-installed with the Open Cowork desktop app under .claude/skills/, or copy the skill folder from the repo into any Agent Skills-compatible environment:

git clone https://github.com/OpenCoworkAI/open-cowork.git
cp -r open-cowork/.claude/skills/skill-creator <your-skills-directory>/

How do you use this skill?

Try saying

Once installed, send your agent any of these to trigger it:

  • Create a skill for handling company expense reports—ask me first what features it should support
  • I keep asking you to rotate PDFs; wrap that repeated operation into a custom skill with a script
  • Build a query skill for my database schema and put the table structures in a references directory
  • Package my skill folder into a .skill file and validate that it meets the requirements

When you express the intent to create or update a skill, the AI follows the six-step flow in SKILL.md: understand examples → plan resources → initialize → edit → package → iterate. The key commands are:

scripts/init_skill.py <skill-name> --path <output-directory>
scripts/package_skill.py <path/to/skill-folder> [output-dir]

The packaging script validates YAML frontmatter, naming conventions, and file organization first; it only produces the .skill zip archive if validation passes.

What are this skill's strengths and limitations?

Pros
  • Complete, actionable six-step process from requirements discovery to packaged distribution
  • Ships with init and package scripts including pre-package validation, reducing errors
  • Progressive-disclosure and conciseness principles directly improve the context efficiency of skills you build
  • MIT-licensed and free with the open-source repository
Limitations
  • Companion scripts require Python, which you must have available
  • Aimed at skill authors rather than end business users, so there is a learning curve
  • No test-suite evidence for the bundled scripts is documented in SKILL.md; quality depends on upstream maintenance

How does this skill compare with similar options?

Side by side with related skills; every score comes from the same FSRS standard.

Skill FS score Stars Last updated License
Skill Creator this page 70 · Strongly recommended ★ 2.2k 12d ago MIT
Create New Skills (Meta-Skill) 62 · Recommended ★ 1.5k 8mo ago —
DeepTutor Skill Creator 57 · Use with care ★ 41k 17d ago Apache-2.0
Skill Creator 48 · Use with care ★ 3.1k 8mo ago MIT
Agent Skill Blueprint 40 · Not recommended ★ 18k 13d ago MIT

The source positions it as one of Open Cowork's five built-in skills (alongside pptx, docx, pdf, and xlsx) and the FAQ names it as the official tool for custom skill creation; no third-party alternatives are mentioned.

How did FollowSkills review this skill?

FollowSkills review · FSRS-1.0
Strongly recommended
70/ 100 5-point scale 3.5 / 5
1Utility15 / 20 · 3.8/5

A bundled copy of Anthropic's official 'how to create a skill' guide — pure methodology content (conciseness principle, progressive disclosure), no execution risk.

2Reliability14 / 20 · 3.5/5
3Safety21 / 25 · 4.2/5
4Evidence6 / 15 · 2.0/5
5Usability8 / 10 · 4.0/5
6Maintenance6 / 10 · 3.0/5
1 2 3 4 5 6

Open a dimension to read why it scored that way

Reviewed Jul 17, 2026 Review evidence[1]

Evidence confidence:Low — Mostly static review, author material or a limited demo; useful for discovery, not high-risk decisions.

This review was completed under FSRS 1.0. Dimension scores are not converted; it is queued for re-review under FSRS 2.0.

See the full review method →

FAQ

Do I need to write code to create a skill?
Not necessarily. The minimum is a SKILL.md (YAML frontmatter plus Markdown instructions); scripts, templates, and other resources are optional, and init_skill.py generates a scaffold to fill in.
What is a .skill file?
The packaging script outputs a .skill file that is simply a zip archive with a .skill extension, containing all skill files with the proper directory structure.
What happens if validation fails?
package_skill.py reports the errors and exits without creating a package; fix the issues and rerun the command.
Does using this skill cost anything?
No. The repo is MIT-licensed and Open Cowork is free; you only pay for underlying AI model API usage from your chosen provider.

More skills from this repository

All from OpenCoworkAI/open-cowork

Related skills