claudekit / tools / last30days
[ Skill · Data Engineering ]

Last30Days

An AI agent skill that researches any topic across Reddit, X, YouTube, HN, Polymarket, and the web — then synthesizes a grounded summary. The engine scores results by actual engagement metrics (upvotes, likes, view counts, prediction market odds) rather than editorial authority, resolves entities first (handles, subreddits, GitHub repos, hashtags), runs parallel multi-source queries, merges duplicate stories across platforms, and produces cited briefs. Reddit, HN, Polymarket, and GitHub work immediately with zero configuration.

mvanhorn/last30days-skill ·published ·updated
$ /plugin marketplace add mvanhorn/last30days-skill copy

A community open-source project by mvanhorn (mvanhorn/last30days-skill); v3 engine by j-sperling.

What it does

Finding out what people are actually saying about a topic means searching Reddit, X, YouTube, and Hacker News separately and synthesizing scattered threads by hand. A single web search surfaces SEO-optimized articles first, while real community signal sinks.

Last30Days collapses that into one command. Given a topic, it resolves entities first — finding relevant handles, subreddits, GitHub repos, and hashtags — before parallel multi-source queries, merges duplicate stories across platforms, and produces cited briefs. Ranking is driven by actual engagement metrics — upvotes, likes, view counts, prediction market odds — rather than editorial authority.

Data sources

SourceHow accessedCost
Reddit (with comments)Public JSON APIFree
Hacker NewsPublic APIFree
PolymarketPublic APIFree
GitHubPublic APIFree
X/TwitterBrowser login (no API key needed)Free
YouTubeyt-dlp (CLI tool)Free
BlueskyApp password from bsky.appFree
TikTok, Instagram, Threads, PinterestScrapeCreators API key100 free credits, then pay-as-you-go
Perplexity SonarOpenRouter API keyPay-as-you-go
Web searchBrave Search API key2,000 free queries/month

Reddit, HN, Polymarket, and GitHub work immediately with zero configuration. The first run triggers a setup wizard for additional sources.

Quick start

  1. Install — /plugin marketplace add mvanhorn/last30days-skill in Claude Code, or npx skills add mvanhorn/last30days-skill -g on other hosts.
  2. Run /last30days <topic> — a person, a product, an event, or a comparison.
/last30days Peter Steinberger
/last30days OpenClaw vs Hermes vs Paperclip
/last30days OpenAI --competitors
/last30days OpenClaw --emit=html
  1. After a run, type eli5 on for a plain-language rewrite; eli5 off reverts.
  2. --emit=html produces a self-contained, dark-mode, offline-compatible brief saved as {topic}-brief.html.

Notes

  • MIT license — community project mvanhorn/last30days-skill.
  • X access uses browser login — instead of an API key, X is accessed through your browser login (a vendored Bird client). Be aware queries run under your own account.
  • Install count source — the ⬇ 9.4k install count on this page comes from skills.sh npx skills telemetry (checked 2026-06-11); it is a different source from claude.com/plugins installs.
  • Dependencies — Python 3.12+, yt-dlp, and Node.js.
  • API key storage — .env file, environment variables, or macOS Keychain via skills/last30days/scripts/setup-keychain.sh.
  • Trend monitoring — the --store flag persists findings to SQLite; scripts/watchlist.py and scripts/briefing.py support scheduled runs and digests.
  • Save location — defaults to ~/Documents/Last30Days/; override with LAST30DAYS_MEMORY_DIR or --save-dir <path>. The README states “No tracking; research stays on your machine.”
§ 5

See also

same category · curated
[01]
[Tool] Firecrawl · 🔥 The API to search, scrape, and interact with the web for AI. Three integrated capabilities — Search, Scrape, Interact — exposed through one API. Open source under AGPL-3.0 and self-hostable via docker-compose, the engine also powers the firecrawl.dev cloud SaaS run by the same team.
tool · claudekit.io / tools / firecrawl
[02]
[MCP] Firecrawl MCP · 🔥 Official Firecrawl MCP Server — Adds powerful web scraping and search to Cursor, Claude and any other LLM clients. Exposes 12+ tools spanning single-page scrape, batch processing, site crawl, search, structured extraction, autonomous research agent, and interactive page automation, returning clean LLM-ready markdown.
tool · claudekit.io / tools / firecrawl-mcp
[03]
[MCP] Perplexity MCP · The official MCP server implementation for the Perplexity API Platform. Provides AI assistants with real-time web search, reasoning, and research capabilities through Sonar models and the Search API.
tool · claudekit.io / tools / perplexity-mcp
§ 6

Frequently Asked Questions

frequently asked
§ 6.1
What is Last30Days?
An AI agent skill that researches any topic across Reddit, X, YouTube, HN, Polymarket, and the web — then synthesizes a grounded summary. Results are scored by actual engagement metrics (upvotes, likes, view counts, prediction market odds) rather than editorial authority.
§ 6.2
Do I need API keys?
Not to start — Reddit, Hacker News, Polymarket, and GitHub work immediately with zero configuration. Optional sources need keys: ScrapeCreators for TikTok/Instagram/Threads/Pinterest, OpenRouter for Perplexity Sonar, and Brave Search for web search (2,000 free queries/month). A setup wizard runs on first launch.
§ 6.3
Which environments does it work with?
Claude Code (plugin marketplace), claude.ai on the web (upload the .skill file), and 50+ Agent Skills hosts including Codex, Cursor, GitHub Copilot, and Gemini CLI via npx skills.
§ 6.4
How do I use it?
Run /last30days followed by any topic — a person, a product comparison (A vs B), or an event. Add --emit=html for a shareable HTML brief, and type eli5 on after a run for a plain-language rewrite.
§ 6.5
Where does the research data go?
Per the README, "research stays on your machine" — briefs save to ~/Documents/Last30Days/ by default (override with LAST30DAYS_MEMORY_DIR), with no tracking.
§ 6.6
Is it free?
Yes — open source under the MIT license. Only optional sources (ScrapeCreators for social video platforms, Perplexity via OpenRouter) carry external pay-as-you-go API costs.