Initial commit

This commit is contained in:
Zhongwei Li
2025-11-30 08:42:03 +08:00
commit e32589a11a
8 changed files with 387 additions and 0 deletions

160
skills/firecrawl/SKILL.md Normal file
View File

@@ -0,0 +1,160 @@
---
name: firecrawl
description: Web scraping, search, and data extraction using Firecrawl API. Use when users need to fetch web content, discover URLs on sites, search the web, or extract structured data from pages.
---
# Firecrawl
## Overview
Firecrawl is a powerful web scraping and search API. This plugin provides a token-efficient interface for Claude Code through slash commands.
## When to Use This Skill
- **Scraping**: Fetch content from a single URL as markdown
- **Mapping**: Discover all URLs on a website
- **Searching**: Search the web and optionally scrape results
- **Extracting**: Pull structured data from pages using LLM
## Quick Reference
| Task | Command |
|------|---------|
| Scrape a page | `/firecrawl:scrape <url>` |
| Map a site | `/firecrawl:map <url>` |
| Search the web | `/firecrawl:search <query>` |
| Extract data | `/firecrawl:extract <url> --prompt "..."` |
## Commands
### Scrape - Get Page Content
Fetches and converts a web page to clean markdown.
```
/firecrawl:scrape https://example.com
/firecrawl:scrape https://example.com --format summary
```
**Options:**
- `--format markdown|summary` - Output format (default: markdown)
**Best for:**
- Reading documentation
- Fetching article content
- Getting page text for analysis
### Map - Discover URLs
Finds all URLs on a website. Useful for understanding site structure.
```
/firecrawl:map https://example.com
/firecrawl:map https://example.com --limit 50
```
**Options:**
- `--limit N` - Maximum URLs to return (default: 100)
**Best for:**
- Understanding site structure
- Finding specific pages before scraping
- Discovering documentation sections
### Search - Web Search
Searches the web and returns results with optional content scraping.
```
/firecrawl:search typescript tutorials
/firecrawl:search "react hooks guide" --limit 10
```
**Options:**
- `--limit N` - Maximum results (default: 5)
**Supports search operators:**
- `"exact phrase"` - Exact match
- `-term` - Exclude term
- `site:example.com` - Limit to domain
- `intitle:word` - Word in title
**Best for:**
- Finding information across the web
- Researching topics
- Finding documentation
### Extract - Structured Data
Extracts specific data from pages using LLM prompts.
```
/firecrawl:extract https://example.com --prompt "Extract the main heading and description"
/firecrawl:extract https://example.com/product --prompt "Extract price and features" --schema '{"price": {"type": "number"}, "features": {"type": "array"}}'
```
**Options:**
- `--prompt "text"` - What to extract (required)
- `--schema '{...}'` - JSON Schema for structured output
**Best for:**
- Extracting specific data points
- Getting structured information
- Pulling prices, names, dates, etc.
## Environment Setup
Requires `FIRECRAWL_API_KEY` environment variable.
Get your API key from: https://firecrawl.dev
## Examples
### Research a Topic
```
/firecrawl:search "best practices for TypeScript error handling" --limit 5
```
### Read Documentation
```
/firecrawl:scrape https://docs.example.com/api/authentication
```
### Map a Documentation Site
```
/firecrawl:map https://docs.example.com --limit 200
```
### Extract Product Info
```
/firecrawl:extract https://store.example.com/product/123 --prompt "Extract product name, price, and availability"
```
## Token Efficiency
This plugin is designed for minimal token consumption:
- **Scrape**: Returns clean markdown, truncated at 8000 chars
- **Map**: Shows up to 50 URLs with titles
- **Search**: Limits to 10 results with summaries
- **Extract**: Returns only requested data
Compare to the full MCP server which loads ~14k tokens just from tool definitions.
## Error Handling
Common errors:
- `API key required` - Set FIRECRAWL_API_KEY
- `Invalid URL` - Check URL format
- `Rate limited` - Wait and retry (auto-handled)
- `Site blocked` - Some sites block scraping
## Pricing Note
Firecrawl charges per operation:
- Scrape: 1 credit per page
- Map: 1 credit per call
- Search: 1 credit per result
- Extract: Varies by complexity
Check https://firecrawl.dev/pricing for current rates.