About community-pulse
community-pulse is a skill for people tracking Claude Code feedback across blogs, forums, and social media. It uses Exa searches to classify sentiment and produce sourced, actionable findings in quick or deep mode.
Key features
- 01Searches five Claude Code feedback categories
- 02Offers quick and deep search modes
- 03Classifies sentiment of results
- 04Reports sourced actionable insights
Use cases
- 01Review recent Claude Code bug reports
- 02Track community reactions to updates and plugins
- 03Gather community tips and comparisons
Install
Copy the command and run it in a terminal.
$
npx skills add https:/ / github.com/ fusengine/ agents --skill community-pulseAfter installing, type/community-pulsein Claude Code to confirm it loaded.
Bundled filesreferences/exa-queries.mdreferences/templates/pulse-report.md
Details
From the project README
Gathers and analyzes community feedback about Claude Code — real-world usage patterns, bug reports, and feature requests from blogs, forums, and social media — using Exa search tools. Two modes: quick (default, `web_search` for recent mentions) or deep (`--deep`, `deep_researcher` for comprehensive analysis). Searches five priority categories (updates, bugs, plugins, comparisons, tips), classifies each result's sentiment (positive/neutral/negative), and reports actionable insights with sources.
Community Pulse Skill
Overview
Gathers and analyzes community feedback about Claude Code using Exa search tools.
Modes
- Quick (default): Exa web_search for recent mentions
- Deep (--deep): Exa deep_researcher for comprehensive analysis
Search Categories
| Category | Exa Query | Priority |
|---|---|---|
| Updates | "Claude Code" release OR update 2026 | HIGH |
| Bugs | "Claude Code" bug OR issue OR broken | HIGH |
| Plugins | "Claude Code" plugin OR hooks experience | MEDIUM |
| Comparison | "Claude Code" vs Cursor OR Windsurf | LOW |
| Tips | "Claude Code" tips OR workflow OR best practices | MEDIUM |
Workflow
- Search using category-specific Exa queries
- Filter for recent results (last 30 days)
- Analyze sentiment (positive/neutral/negative)
- Extract actionable insights
- Report with sources and recommendations
Sentiment Classification
| Signal | Classification |
|---|---|
| "love", "amazing", "game changer" | Positive |
| "update", "released", "changed" | Neutral |
| "broken", "bug", "regression", "worse" | Negative |
References
Comments
Markdown supported