Claude Code Plugins

Community-maintained marketplace

Feedback

On-demand trend scan for a topic. Search HN and GitHub for what's happening now.

Install Skill

1Download skill
2Enable skills in Claude

Open claude.ai/settings/capabilities and find the "Skills" section

3Upload to Claude

Click "Upload skill" and select the downloaded ZIP file

Note: Please verify skill by going through its instructions before using it.

SKILL.md

name scan
description On-demand trend scan for a topic. Search HN and GitHub for what's happening now.
allowed-tools mcp__pattern-radar__scan_trends, mcp__pattern-radar__explore_pattern, mcp__perplexity-search__perplexity_search

Trend Scanning

Scan current trends across Hacker News and GitHub for any topic.

When to Use

  • "What's happening in [domain]?"
  • Research before starting a new project
  • Staying current with technology trends
  • Finding emerging tools and patterns

Basic Scan

scan_trends(topic: "WebAssembly")

Returns:

  • Signals from HN and GitHub
  • Detected patterns across sources
  • Relevance scoring based on your profile
  • Actionable suggestions

Targeted Scans

HN only (community discussions):

scan_trends(topic: "AI agents", sources: ["hackernews"])

GitHub only (code and repos):

scan_trends(topic: "rust CLI tools", sources: ["github"])

Deep Dive

After finding interesting signals:

explore_pattern(topic: "specific-pattern")

Perplexity Integration

For aggregated web search across Reddit, Twitter, arXiv, and more:

perplexity_search("your topic in the context of your domains")

Pattern-radar generates ready-to-use Perplexity queries in scan results.

Understanding Results

Signals: Individual items (HN stories, GitHub repos)

  • Score: Engagement level (HN points, GitHub stars)
  • Source: Where it came from

Patterns: Clusters of related signals

  • Relevance: How well it matches your domains
  • Actions: Suggested next steps

Tips

  1. Scan broad topics first, then narrow down
  2. Use your profile domains for automatic relevance scoring
  3. Combine with perplexity_search for comprehensive coverage
  4. Save interesting patterns to shared-memory