banana-skill-finder

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Banana Skill Finder

Banana技能查找器

Proactively helps users discover and install relevant Claude skills when they encounter tasks that could benefit from specialized capabilities.

当用户遇到可借助专业能力完成的任务时，主动帮助用户发现并安装相关的Claude技能。

When to Use This Skill

何时使用此技能

Trigger automatically (without user request) when detecting:

Working with specific file formats or technologies
Describing repetitive or specialized tasks
Asking "is there a skill/tool for..." or similar
Struggling with domain-specific work
Any task where a specialized skill could help

Important: This skill should trigger proactively. Don't wait for users to explicitly ask for skill recommendations.

当检测到以下情况时，自动触发（无需用户请求）：

处理特定文件格式或技术
描述重复性或专业性任务
询问“是否有针对……的技能/工具”或类似问题
在特定领域工作中遇到困难
任何专业技能可提供帮助的任务

重要提示：此技能应主动触发，不要等待用户明确请求技能推荐。

Workflow

工作流程

1. Analyze User Need

1. 分析用户需求

Identify:

Core task: What is the user trying to accomplish?
Domain: What category does this fall into? (development, documents, data, web, devops, content, etc.)
Keywords: Extract 2-4 relevant search terms

确定：

核心任务：用户想要完成什么？
领域：属于哪个类别？（开发、文档、数据、Web、DevOps、内容等）
关键词：提取2-4个相关搜索词

2. Search for Skills

2. 搜索技能

Use a three-tier strategy with automatic fallback:

Tier 1: SkillsMP API (Best - if configured)

bash

undefined

采用三层策略，自动降级备选：

第一层：SkillsMP API（最佳选择 - 若已配置）

bash

undefined

Check for API key

echo $SKILLSMP_API_KEY

If exists, use AI semantic search

curl -X GET "https://skillsmp.com/api/v1/skills/ai-search?q={natural_language_query}"
-H "Authorization: Bearer $SKILLSMP_API_KEY"


Benefits:
- AI understands user intent, not just keywords
- Access to 60,000+ curated skills
- Best relevance and quality indicators

**Tier 2: skills.sh WebFetch (Good - always works)**
```bash

curl -X GET "https://skillsmp.com/api/v1/skills/ai-search?q={natural_language_query}"
-H "Authorization: Bearer $SKILLSMP_API_KEY"


优势：
- AI能理解用户意图，而非仅匹配关键词
- 可访问60,000+经过筛选的技能
- 最佳的相关性和质量指标

**第二层：skills.sh WebFetch（良好选择 - 始终可用）**
```bash

Try search with query parameter

Use WebFetch: https://skills.sh/?q={keywords}

Or browse leaderboard

Use WebFetch: https://skills.sh # All-time popular Use WebFetch: https://skills.sh/trending # Trending (24h)


Benefits:
- 200+ high-quality curated skills
- No authentication needed
- Ranked by install count
- Shows trending skills

**Tier 3: GitHub API (Fallback - may have limits)**
```bash
curl -X GET "https://api.github.com/search/code?q={keywords}+SKILL.md+language:markdown" \
  -H "Accept: application/vnd.github.v3+json"

Note: Rate limited (60/hour unauthenticated), use only as last resort.

Optional: Check Local Installed Skills

bash

ls ~/.claude/skills/

Check if user already has relevant skills installed but hasn't used them.

Recommendation Order: Try Tier 1 → Tier 2 → Tier 3. Stop when you find good matches.

Use WebFetch: https://skills.sh # All-time popular Use WebFetch: https://skills.sh/trending # Trending (24h)


优势：
- 200+高质量筛选后的技能
- 无需身份验证
- 按安装量排名
- 展示热门技能

**第三层：GitHub API（备选方案 - 可能有限制）**
```bash
curl -X GET "https://api.github.com/search/code?q={keywords}+SKILL.md+language:markdown" \
  -H "Accept: application/vnd.github.v3+json"

注意：有调用频率限制（未验证用户每小时60次），仅作为最后手段使用。

可选：检查本地已安装技能

bash

ls ~/.claude/skills/

检查用户是否已安装相关技能但尚未使用。

推荐顺序：尝试第一层 → 第二层 → 第三层。找到合适匹配项后即停止。

3. Rank by Relevance

3. 按相关性排名

Score each found skill based on:

Keyword match with user's need (most important)
Functionality alignment
Quality indicators (stars, recent activity)
Specificity vs generality

Select the 1-3 most relevant skills. Quality over quantity.

根据以下标准为每个找到的技能评分：

与用户需求的关键词匹配度（最重要）
功能契合度
质量指标（星标数、近期活跃度）
特异性与通用性

选择1-3个最相关的技能。质量优先于数量。

4. Present Recommendations

4. 展示推荐结果

Format recommendations as:

I found [N] skill(s) that could help:

**1. [Skill Name]** - [One-line description]
   Source: [SkillsMP/GitHub/Vercel]
   Repository: [owner/repo]
   Why relevant: [Brief explanation]
   Install: `npx skills add [owner]/[repo]`

[Repeat for 2-3 skills max]

Would you like me to install any of these?

推荐结果格式如下：

I found [N] skill(s) that could help:

**1. [Skill Name]** - [One-line description]
   Source: [SkillsMP/GitHub/Vercel]
   Repository: [owner/repo]
   Why relevant: [Brief explanation]
   Install: `npx skills add [owner]/[repo]`

[Repeat for 2-3 skills max]

Would you like me to install any of these?

5. Install if Approved

5. 获许后安装

When user approves, install using Vercel's skills CLI:

bash

npx skills add <owner>/<repo>

Examples:

bash

npx skills add vercel-labs/agent-skills
npx skills add anthropics/skills

This command:

Downloads the skill from GitHub
Installs to
```
~/.claude/skills/
```
Works with Claude Code, Cursor, Windsurf, and other agents
Tracks installation via anonymous telemetry (leaderboard)

Confirm installation success and explain how the skill will help.

当用户同意后，使用Vercel的skills CLI进行安装：

bash

npx skills add <owner>/<repo>

示例：

bash

npx skills add vercel-labs/agent-skills
npx skills add anthropics/skills

此命令：

从GitHub下载技能
安装到
```
~/.claude/skills/
```
目录
可与Claude Code、Cursor、Windsurf及其他agent配合使用
通过匿名遥测跟踪安装情况（排行榜）

确认安装成功，并说明该技能将如何提供帮助。

Key Principles

核心原则

Proactive, Not Reactive: Trigger automatically when relevant, don't wait to be asked
Quality Over Quantity: Recommend only 1-3 best matches, not a long list
Smart Three-Tier Search:
- Tier 1: SkillsMP AI search (best, if configured)
- Tier 2: skills.sh leaderboard (good, always works)
- Tier 3: GitHub API (fallback, rate limited)
- Stop when you find good matches
Explain Relevance: Always explain why each skill matches their need
Easy Installation: Use
```
npx skills add owner/repo
```
for one-command installation
API Key Recommended but Optional: Best results with SkillsMP API key, but skills.sh fallback works well

主动触发，而非被动响应：相关时自动触发，不要等待用户请求
质量优先于数量：仅推荐1-3个最佳匹配项，而非冗长列表
智能三层搜索：
- 第一层：SkillsMP AI搜索（最佳，若已配置）
- 第二层：skills.sh排行榜（良好，始终可用）
- 第三层：GitHub API（备选，有调用限制）
- 找到合适匹配项后即停止
说明相关性：始终解释每个技能为何匹配用户需求
安装便捷：使用
```
npx skills add owner/repo
```
实现一键安装
API密钥推荐但非必需：配置SkillsMP API密钥可获得最佳结果，但skills.sh备选方案在大多数情况下表现良好

Examples

示例

User says: "I need to extract text from a PDF file" → Trigger skill-finder, search for PDF processing skills, recommend pdf-editor or similar

User says: "Help me review this React component" → Trigger skill-finder, search for React/code-review skills, recommend react-best-practices from Vercel

User says: "I'm deploying to AWS" → Trigger skill-finder, search for AWS/deployment skills, recommend cloud-deploy or aws-helper

User says: "How do I query this BigQuery table?" → Trigger skill-finder, search for BigQuery/SQL skills, recommend bigquery or data-analysis skills

用户说："我需要从PDF文件中提取文本" → 触发技能查找器，搜索PDF处理技能，推荐pdf-editor或类似技能

用户说："帮我审查这个React组件" → 触发技能查找器，搜索React/代码审查技能，推荐Vercel的react-best-practices

用户说："我正在部署到AWS" → 触发技能查找器，搜索AWS/部署技能，推荐cloud-deploy或aws-helper

用户说："我该如何查询这个BigQuery表？" → 触发技能查找器，搜索BigQuery/SQL技能，推荐bigquery或数据分析技能

Additional Resources

其他资源

For detailed information:

references/api_config.md - How to set up SkillsMP API key
references/skill_sources.md - Skill sources, categories, and search strategies

如需详细信息：

references/api_config.md - 如何设置SkillsMP API密钥
references/skill_sources.md - 技能来源、类别及搜索策略

Setup Recommendations

设置建议

For best results, suggest users configure SkillsMP API key:

Visit https://skillsmp.com/docs/api
Generate API key
Set environment variable:
```
export SKILLSMP_API_KEY="sk_live_..."
```

This enables AI semantic search (much better than keyword matching). Without it, the skill automatically falls back to skills.sh leaderboard search, which still works well for most cases.

为获得最佳效果，建议用户配置SkillsMP API密钥：

访问https://skillsmp.com/docs/api
生成API密钥
设置环境变量：
```
export SKILLSMP_API_KEY="sk_live_..."
```

这将启用AI语义搜索（远优于关键词匹配）。若未配置，该技能会自动降级使用skills.sh排行榜搜索，在大多数情况下仍能表现良好。