last30days-skill
SHA-256An AI agent skill that searches Reddit, X, YouTube, HN, Polymarket and the web, then synthesizes a grounded summary based on real engagement signals (upvotes, likes, money bets). Beat Google's stale SEO by tapping what people actually talk about.
Smart Download
Download Download Version
v3.11.0 · 5.2 MB
Cross-platform social search engine powered by AI, scoring content by real engagement, not algorithms
Core Features
- Simultaneously search Reddit, X, YouTube, TikTok, HN, Polymarket, GitHub, and more
- Rank results by real engagement (upvotes, likes, bets) instead of editorial curation
- AI judge auto-generates a cited synthesis, supporting shareable HTML briefs
- Zero config for Reddit/HN/GitHub; optional wizard unlocks X/YouTube/TikTok in 30s
- Compatible with 50+ AI hosts including Claude Code, Cursor, Copilot, Gemini CLI
What It Can't Do
- •Some sources (X, YouTube, TikTok) require manual API key or browser session setup
- •Search is limited to the last 30 days; cannot go further back
- •Summary quality depends on the AI host model; may have AI bias
- •Free to use, but heavy usage may hit API rate limits of individual platforms
Use Cases
- Pre-meeting research: learn what a CEO or partner has actually been doing in the last 30 days
- Trend tracking: get real-time heatmap of a topic across multiple social platforms
- Tool comparison: compare open-source projects or products using GitHub stars, community threads, and real-time reviews
Detailed Introduction
Last30days is a powerful AI agent skill that turns your favorite AI assistant (Claude Code, Cursor, Copilot, etc.) into a multi-platform search engine. Unlike Google or ChatGPT which are limited by API deals, this skill simultaneously queries Reddit, X/Twitter, YouTube, TikTok, Hacker News, Polymarket, GitHub, and more—scoring results by real human engagement rather than editorial curation. It then uses a judge AI to produce a concise, cited summary. Zero config for Reddit/HN/GitHub; optional setup wizard unlocks X, YouTube, TikTok in 30 seconds. Perfect for competitive research, pre-meeting briefs, trend analysis, and comparing tools or people. Compared to traditional search, it surfaces the content that actually moved communities, not SEO-optimized blogs.
Tags
Getting Started
Install the software
Double-click the downloaded installer and follow the prompts
If using Claude Code, run the plugin marketplace command to install
Or use npx command to install globally (works with most AI hosts)
Type /last30days [topic] to start searching. Reddit/HN/GitHub work immediately
- If using Claude Code, run the plugin marketplace command to install
- Or use npx command to install globally (works with most AI hosts)
- Type /last30days [topic] to start searching. Reddit/HN/GitHub work immediately
SHA-256 checksum verified
Checksum extracted from GitHub official Release page
SHA256 Checksum
d0fbae6ab358f6f14a998b447cb252579935dad795d9fddd6205f9fee689a114This checksum is extracted from the GitHub Release page. Verify file integrity after download.
All SHA-256 checksums on this platform are extracted from the project's official GitHub Release page, without any modification. You can independently verify them on the GitHub Releases page.
Open Source Transparency
View GitHub SourceUninstall Info
For Claude Code: run /plugin uninstall last30days. For npx installations: run npx skills remove mvanhorn/last30days-skill -g to uninstall globally.
No Extra Dependencies
Ready to use after download. No additional runtime required.
Similar Projects
goose
Goose is a general-purpose AI agent that runs on your machine. Use it for coding, research, writing, automation, and more. It offers a native desktop app (macOS, Linux, Windows), a full CLI, and an API. Works with 15+ providers (Anthropic, OpenAI, Google, Ollama, etc.) and connects to 70+ extensions via the Model Context Protocol. Built in Rust, part of the Linux Foundation.
ollama
Ollama lets you download, run, and manage large language models locally. One command, multiple platforms, endless possibilities.
llama.cpp
High-performance LLM inference engine in C/C++ with minimal dependencies, supporting quantized models (1.5–8 bit) and diverse hardware (Apple Silicon, CUDA, Vulkan, etc.).