# 用于网页抓取和 Markdown 提取的 Firecrawl CLI
把网页提取为 Markdown 或结构化数据，适合 Agent 研究和检索流程。
Agent Readiness Score: 76/100
Good for approved web extraction, but crawling scope and API keys need clear limits.
## Install

- npx: `npx firecrawl`

## Structured output

Supported output formats: markdown, json.
Prefer structured output flags such as `--json`, `--format json`, or equivalent when available.

## Safe commands

- Scrape one URL: `npx firecrawl scrape https://example.com` — Extracts a single page before broader crawling.

## Commands requiring approval

- Crawl a page: `npx firecrawl crawl https://example.com` — requires-review; require explicit user approval.

## Agent instructions

- Codex CLI: Use Firecrawl for a small list of approved URLs, preserve source links, and summarize extracted markdown separately from your conclusions.

## Safety notes

- Respect robots, rate limits, and site terms before crawling.
- Agents should crawl a small target set first and store source URLs with extracted content.

## Metadata

Official URL: https://www.firecrawl.dev/
GitHub: https://github.com/mendableai/firecrawl
Docs: https://docs.firecrawl.dev/