Data & Research
Open datasets, export formats, and citation info for researchers.
Available Datasets
Three public datasets cover the full catalog. All figures come from live database queries — nothing is estimated or fabricated.
- skills — every indexed skill with installs, GitHub stars, flags, and source repo
- prompts — AI image-generation prompts with text, model, author, and engagement counts
- organizations — skill-publishing organizations with followers and skill counts
Accessing the Data
Use the REST API for quick summaries, or the export endpoint for full dataset downloads (max 10,000 rows per request).
Statistics summary (JSON)
curl "https://expliki.org/api/v1/stats"Full dataset export (JSON or CSV)
curl "https://expliki.org/api/v1/export?type=skills&format=csv" -o skills.csv
curl "https://expliki.org/api/v1/export?type=prompts&format=json" -o prompts.json
curl "https://expliki.org/api/v1/export?type=organizations&format=csv" -o organizations.csvPython
import json, urllib.request
url = "https://expliki.org/api/v1/export?type=skills&format=json"
with urllib.request.urlopen(url) as r:
dataset = json.load(r)
for skill in dataset["data"]:
print(skill["slug"], skill["installs"])Exports are limited to 10,000 rows per request and 10 requests per minute per IP. All endpoints are CORS-enabled and need no API key.
Methodology
Skills data is aggregated from the public skills.sh leaderboard and the GitHub API; prompts from public source sitemaps. Rankings use install counts only — no invented ratings.
License
The datasets are released under CC BY 4.0
Cite This Dataset
If you use this data in research or journalism, please cite it as:
APA
Git Codar. (2026). Git Codar: AI Agent Skills & Prompts Dataset [Data set]. https://expliki.org/statsBibTeX
@misc{gitcodar2026,
author = {{Git Codar}},
title = {Git Codar: AI Agent Skills \& Prompts Dataset},
year = {2026},
howpublished = {\url{https://expliki.org/stats}},
note = {Accessed: 2026-07-21},
license = {CC BY 4.0}
}