The CLI docs do a good job separating the web work modes — scrape for clean Markdown from specific URLs, interact when clicks/forms are needed, and agent for longer structured extraction jobs.
One small docs addition that might help new users: an optional quick-try link for the firecrawl-scrape Skill, so they can test the Skill inside an OpenClaw/Hermes-style agent before setting up the CLI locally:
Try Firecrawl Web Scraping Skill in Telegram or WhatsApp
It would just be a low-friction first run next to the existing install/auth docs, not a replacement for the local Firecrawl setup.
The CLI docs do a good job separating the web work modes —
scrapefor clean Markdown from specific URLs,interactwhen clicks/forms are needed, andagentfor longer structured extraction jobs.One small docs addition that might help new users: an optional quick-try link for the
firecrawl-scrapeSkill, so they can test the Skill inside an OpenClaw/Hermes-style agent before setting up the CLI locally:Try Firecrawl Web Scraping Skill in Telegram or WhatsApp
It would just be a low-friction first run next to the existing install/auth docs, not a replacement for the local Firecrawl setup.