agent-skills-fetcher

# Agent-Skills.md Skill Fetcher Fetches the complete skill catalog from agent-skills.md and writes it to a local text file. ## What it does - Paginates through `https://agent-skills.md/?page={page}` (pages 1-1000) - Parses HTML `<main>` content to extract skill cards - Each `<a>` card contains: `<h3>` (name), `<p>` (description), star `<span>` (stars) - Extracts: `name`, `subtitle` (empty), `brief`, `category` (empty), `url`, `stars` - De-duplicates by full URL - Outputs one skill per line, tab-separated, to `agent-skills.txt` ## Usage Run the helper script from the skill directory: ```bash node .trae/skills/agent-skills-fetcher/scripts/fetch.js ``` The script will: 1. Iterate pages 1 through 1000 2. Extract skill cards from `<main>` HTML 3. De-duplicate by full URL 4. Write results to `agent-skills.txt` in the current working directory 5. Stop early if a page returns 0 skills ## Output format ``` name\tsubtitle\tbrief\tcategory\turl\tstars ``` Example line: ``` github\t\tGitHub operations via `gh` CLI: issues, PRs, CI runs...\thttps://agent-skills.md/skills/steipete/clawdis/github\t341,754 ``` ## Parameters - `maxPage` — last page to fetch (default: 1000) - `outputFile` — destination path (default: `./agent-skills.txt`) - `delayMs` — delay between pages in ms (default: 1500)