Hot skills
Agent Skills 技能集合，将各领域经验打包为可复用的 AI 能力 | A collection of Agent Skills that turn AI agents into reusable domain experts.
npx -y skills add xiehuacheng/skills --skill hot-skillsAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- 24 days oldThe repository was created 24 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
- 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Discover hot and trending Agent Skills across multiple leaderboards (agentskills.media, skills-rank.com, skills.sh). Summarize, categorize, and search. Use when the user asks about popular skills, hot skills, trending skills, or wants to find skills by category.
SKILL.md
5.3 KB, as published. Nobody here has run it
Hot Skills
This skill helps you discover trending and hot Agent Skills by aggregating data from multiple sources:
- agentskills.media — GitHub stars and categories
- skills-rank.com — Individual skill ranking scores
- skills.sh — Real install counts (fetched via headless browser from the public leaderboard)
Boundaries and Defaults
Can do: aggregate and de-duplicate data from three public leaderboards; summarize top skills overall or by category; search by keyword; output JSON or human-readable text; force-refresh local cache; filter by category or alias.
Cannot do (without explicit user approval): install or modify any Agent Skill on the user's system; write to files outside the user's chosen --output directory; assume default categories or aliases without confirmation.
Default behavior: read-only — never writes unless --output <dir> is explicitly passed; uses cached data when fresh (<1h old) to avoid rate-limit issues; shows the data source and freshness on each command.
When to Use
Use this skill when the user:
- Asks "现在有什么热门 skill？"
- Wants to find popular skills in a domain: "前端热门 skill"
- Asks "帮我搜一下 xxx 相关的 skill"
How to Use
Run the main script from the skill directory:
node scripts/fetch-trends.js [options]
When this skill is installed via npx skills add, the skill directory is available at .claude/skills/skills-trending/ (or the equivalent for your agent). Adjust the path accordingly.
Options
| Option | Description | Example |
|---|---|---|
--top, -n | Number of top skills to show | --top 10 |
--category, -c | Filter by category | --category frontend |
--search, -s | Search by keyword | --search "react testing" |
--trending, -t | Show skills.sh 24h trending (Playwright) | --trending --top 10 |
--refresh, -r | Force refresh data | --refresh |
--json, -j | Output JSON | --json |
Examples
# Top 20 hot skills overall
node scripts/fetch-trends.js
# Top 10 frontend skills
node scripts/fetch-trends.js --category frontend --top 10
# Search for testing skills
node scripts/fetch-trends.js --search testing
# Force refresh and output JSON
node scripts/fetch-trends.js --refresh --json
# Show skills.sh 24h trending
node scripts/fetch-trends.js --trending --top 10
Output Format
The script outputs markdown tables by default. Present them cleanly to the user.
For each skill, include:
- Name
- Stars (from agentskills.media)
- Installs (from skills-rank.com / skills.sh)
- Hot score (combined ranking)
- Category
- One-line description
Each report ends with a quick install hint. You can install any listed skill with:
npx skills add <Skill>
Replace <Skill> with the value shown in the Skill column (e.g. obra/superpowers@superpowers).
Common Categories
| Category | Aliases | Typical Queries |
|---|---|---|
| Web Dev | frontend, web, webdev, react, nextjs, typescript, css, tailwind | Frontend frameworks, full-stack web skills |
| UI/UX | ui, ux, design, design-system, accessibility | Design systems, visual design, accessibility |
| Testing | testing, test, jest, playwright, e2e | Unit tests, e2e tests, QA |
| DevOps | devops, docker, kubernetes, k8s, deploy, ci-cd | Deployment, containers, CI/CD |
| Docs | doc, docs, documentation, api-docs, changelog, readme | README, changelog, documentation generation |
| Code Quality | review, lint, refactor, best-practices, code-quality | PR review, linting, refactoring |
| Automation | automation, workflow, git, productivity | Git workflows, task automation |
| AI Agents | agent, agents, memory | Agent frameworks, memory, planning |
| Knowledge | knowledge | Knowledge management, RAG |
| Prompts | prompt, prompts | Prompt engineering, optimization |
| Security | security | Security review, hardening |
| Data | data | Data analysis, visualization |
| Integrations | integration, integrations | Third-party tool integrations |
| Search | search | Search capabilities |
| Planning | planning | Task planning, project management |
| Game Dev | game, gamedev | Game development |
| Writing | writing | Writing, content generation |
| Code Gen | code, codegen | Code generation |
| Awesome List | awesome | Curated awesome lists |
| General | general | General-purpose skills |
Tips
- If data seems stale, add
--refresh. - Category filtering supports partial matches and aliases (e.g.,
--category frontendmatches "Web Dev" and "UI/UX"). - Use specific keywords when searching:
--search "react testing"works better than--search testing. - The first
--refreshmay take 30–60 seconds because data is fetched from several upstream sources (including a headless browser pass for skills.sh). --trendingand the default top-skills fetch both require the Python/Playwright environment for skills.sh data.