Trending Crawler

Trending Crawler

Discovers new AI Agent Skills from GitHub Trending pages and adds them to skills.json.

Workflow

1. Scrape GitHub Trending

cd /tmp && \
curl -s "https://github.com/trending?since=daily" | \
python3 
repo_pattern = r'\s*]*>'
desc_pattern = r']*>([^'
star_pattern = r'aria-label="(\d[\d,]+) star'

repo_matches = re.findall(repo_pattern, html)
desc_matches = re.findall(desc_pattern, html)
star_matches = re.findall(star_pattern, html)

print(f"Repos found: {len(repo_matches)}")
print(f"Descs found: {len(desc_matches)}")
print(f"Stars found: {len(star_matches)}")

# Keywords that indicate a skill-related repo
SKILL_KEYWORDS = [
    'skill', 'agent', 'claude', 'openai', 'mcp', 'tool',
    'prompt', 'workflow', 'task', 'automation', 'cli',
    'awesome', 'framework', 'hub', 'marketplace'
]

results = []
for i, repo in enumerate(repo_matches[:20]):
    desc = desc_matches[i].strip() if i /dev/null && \
git status && \
echo "---" && \
read -p "Commit and push? (y/n) " ans && \
if [ "$ans" = "y" ]; then \
    git commit -m "chore: discover new skills $(date +%Y-%m-%d)" && \
    git push && \
    echo "Done!"; \
fi

Notes

• Always review new skills before adding (check description quality)
• SKILL.md search finds actual skills, trending finds skill-related projects
• Run this 1-2x per week to keep collection fresh
• Source field: "trending" or "skillmd" to track discovery method

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: vivy-yi
• Source: vivy-yi/awesome-skills
• License: MIT

Install and usage instructions live in the source repository linked above.