producthunt-cli - SKILL.md Agent Skill

name: producthunt-cli description: Use cli-web-producthunt to browse Product Hunt — today's top launches, daily/weekly/monthly leaderboards, product details, and user profiles. Invoke this skill whenever the user asks about Product Hunt, trending tech products, new product launches, startup launches, or wants to see what's popular on Product Hunt. Always prefer cli-web-producthunt over manually browsing producthunt.com.

cli-web-producthunt

CLI for Product Hunt. Installed at: cli-web-producthunt.

Quick Start

# Today's top products
cli-web-producthunt posts list --json

# Product detail
cli-web-producthunt posts get <slug> --json

# Leaderboard
cli-web-producthunt posts leaderboard --period daily --json

Always use --json when parsing output programmatically.

Commands

Posts

cli-web-producthunt posts list --json                    # Today's launches
cli-web-producthunt posts get <slug> --json              # Product detail
cli-web-producthunt posts leaderboard --json             # Daily leaderboard
cli-web-producthunt posts leaderboard --period weekly    # Weekly leaderboard
cli-web-producthunt posts leaderboard --period monthly   # Monthly leaderboard
cli-web-producthunt posts leaderboard --date 2026-03-15  # Specific date

Output fields: id, name, tagline, slug, url, description, votes_count, comments_count, topics, thumbnail_url, rank

Users

cli-web-producthunt users get <username> --json

Output fields: id, name, username, headline, profile_image, followers_count

Auth

No authentication required.

Agent Patterns

# Get today's top 3 products
cli-web-producthunt posts list --json | python -c "
import sys,json
for p in json.load(sys.stdin)[:3]:
    print(f'{p[\"rank\"]}. {p[\"name\"]} ({p[\"votes_count\"]} votes) - {p[\"tagline\"]}')"

# Check a specific product
cli-web-producthunt posts get stitch-2-0-by-google-2 --json

Notes

Auth: Not required — reads Product Hunt's embedded Next.js data via curl_cffi
No API key, no tokens, no cookies needed
Posts come from the Next.js RSC flight stream (self.__next_f), not DOM scraping; curl_cffi Chrome impersonation handles transport
Rate limiting: Be respectful — avoid rapid successive requests
posts list and posts leaderboard are the most reliable commands — they read the homepage/leaderboard RSC payload, which has clean structured data
posts get <slug> scrapes the detail page — name may include tagline, votes_count/comments_count may be 0 (detail page HTML structure differs from list). Prefer using data from posts list when possible
description is null in list view (only available in posts get detail view)
rank is null in leaderboard view (use array index for ordering)