A curated collection of AI skills maintained by onData. Skills are tool-agnostic instructions that help AI assistants handle specific tasks consistently — without re-explaining your workflows every time.

A skill is a set of instructions that tells an AI assistant how to handle a specific task: which tools to use, in what order, and how to format the output. Once installed, it activates automatically when relevant.

For example, install the openalex skill and ask $openalex: Find recent papers on urban heat islands with open access PDFs — you'll get a structured list of works with titles, authors, DOIs, and direct download links.

The project follows an open source model: public repo, open contributions, organic growth.

Skills follow the Agent Skills open standard. If you're new to skills, start with a single one to get familiar with how they work:

npx skills add ondata/skills --skill openalex

Once installed, open your AI tool and try it: use $openalex skill to find recent papers on urban heat islands with open access PDFs — you'll get a list of works with titles, authors, DOIs, and direct PDF links.

Once comfortable, you can install the full collection:

npx skills add ondata/skills

Or install manually:

1. Clone the repo: git clone https://github.com/ondata/skills
2. Copy the skill folder (e.g. skills/openalex/) into your AI tool's skills directory
3. Enable the skill in your tool's settings

The npx skills add commands above use skills.sh, a shell tool and currently the most convenient way to install and manage Agent Skills across AI tools.

Skills work best with CLI-based AI tools (Claude Code, Gemini CLI, OpenAI Codex, etc.) — the command-line context gives the agent full access to your environment and makes skill workflows faster and richer.

During installation you'll be asked:

• Which AI tools to make the skill available in (e.g. Claude code, OpenAI Codex, Gemini cli, Cursor, Windsurf, ecc.).
• User or project scope — user skills are available in every project on your machine; project skills are installed into the current repository and shared with anyone who clones it.
• Installation method — Symlink (recommended) creates a single canonical copy that all your agents share; run npx skills update to update. Copy creates independent copies per agent and is useful only when symlinks are not supported on your system.

Each skill has a test battery in evals/ to verify trigger behavior, process correctness, and output quality.

See evals/README.md for the full dashboard and instructions on how to run or contribute evals.

Each skill is a folder inside skills/ with the following layout:

skills/
└── your-skill-name/          ← kebab-case, matches the skill name
    ├── SKILL.md              ← required — instructions + YAML frontmatter
    ├── scripts/              ← optional — executable scripts (Python, Bash…)
    ├── references/           ← optional — supporting documentation
    └── assets/               ← optional — templates, fonts, icons

Every skill starts with a YAML frontmatter block:

---
name: your-skill-name
description: What it does and when to use it. Use when the user says "[trigger phrase]".
---

# Your Skill Name

## Instructions
…

Rules:

• name: kebab-case only, no spaces, no capitals — must match the folder name
• description: required — include both what the skill does and when to trigger it; max 1024 characters
• No README.md inside the skill folder — all documentation goes in SKILL.md or references/
• File must be named exactly SKILL.md (case-sensitive)

Have a skill you find useful? Open a pull request.

1. Fork the repo
2. Create your skill folder under skills/ following the structure above
3. Make sure SKILL.md has valid YAML frontmatter with name and description
4. Add your skill to the catalog table in this README
5. Open a PR with a short description of what the skill does

No formal review process yet — we'll figure it out as we grow.