ToolsIntegrations
firecrawl
Firecrawl — scrape a URL or crawl a site, returning clean markdown. The workhorse for RAG ingestion from the open web.
import { firecrawl } from '@agentskit/tools'
const runtime = createRuntime({
adapter,
tools: [...firecrawl({ apiKey: process.env.FIRECRAWL_API_KEY! })],
})Sub-tools
| Name | Purpose |
|---|---|
firecrawlScrape | One-shot scrape of a single URL |
firecrawlCrawl | Recursive crawl with include/exclude patterns |
Bundled: firecrawl(config).
Config
type FirecrawlConfig = {
apiKey: string
baseUrl?: string // self-hosted Firecrawl
fetch?: typeof fetch
}Example — on-demand RAG
const runtime = createRuntime({
adapter,
tools: [...firecrawl({ apiKey }), ...rag.tools],
})
await runtime.run('Fetch https://example.com/changelog and answer what shipped this week.')Comparison
| Tool | When to use |
|---|---|
| firecrawl | Structured markdown + crawl trees, API-backed, paid |
| reader | Jina Reader — fast, free, single-URL |
| fetchUrl | Zero-dep HTTP GET, no cleanup |
| browserAgent | Full JS-rendered interaction |