A personal collection of agent skills created by wavever.
This repository is the home for my reusable SKILL.md instruction bundles. It follows the same high-level shape as repositories such as mattpocock/skills: a small repository root plus a skills/ directory where each skill is self-contained.
| Skill | Description | Path |
|---|---|---|
ready-to-opensource-skills |
Prepare, audit, and polish a software project so its repository is ready for a safe, professional open-source release. | skills/ready-to-opensource-skills |
/plugin marketplace add wavever/skills
/plugin install skills@wavever
This installs every skill listed in .claude-plugin/plugin.json.
Copy a single skill directory into your agent's skills folder. For example:
npx degit wavever/skills/skills/ready-to-opensource-skills \
~/.claude/skills/ready-to-opensource-skillsIf the destination skills directory did not exist when your agent started, restart the agent once so the new skill is discovered.
Invoke a skill by name, for example:
/ready-to-opensource-skills
You can also describe the task naturally and let the agent trigger the matching skill.
skills/
├── .claude-plugin/
│ ├── marketplace.json
│ └── plugin.json
├── skills/
│ └── ready-to-opensource-skills/
│ ├── SKILL.md
│ └── references/
├── .github/
├── CODE_OF_CONDUCT.md
├── CONTRIBUTING.md
├── LICENSE
├── README.md
├── SECURITY.md
└── package.json
Each skill folder should be self-contained:
skills/<skill-name>/
├── SKILL.md
├── references/ # optional
├── scripts/ # optional
└── assets/ # optional
- Create a new folder under
skills/<skill-name>/. - Add
SKILL.mdwithnameanddescriptionfrontmatter. - Keep supporting resources inside that skill folder (
references/,scripts/,assets/). - Add the skill path to
.claude-plugin/plugin.json. - Add the skill to the table in this README.
- Validate the skill before committing.
A skill name should be lowercase hyphen-case and should match its folder name.
Most skills in this repository are instruction bundles, not applications. Still, always review a skill before installing it, especially if it includes scripts or commands that an agent may execute.
Issues and pull requests are welcome. See CONTRIBUTING.md and the Code of Conduct. For security concerns, see SECURITY.md.
MIT © 2026 wavever