docs(scrape): add map tool as fallback step for SPA/JS-heavy pages

[firecrawl-spring]

Summary

• Adds step 3 in the scrape tool's SPA handling guidance to use firecrawl_map with a search parameter before falling back to firecrawl_agent
• This helps agents discover the correct page URL on large documentation sites where content is spread across multiple pages

Context

When scraping JavaScript-heavy pages like API documentation sites, the scrape tool may return empty or minimal content. Previously, the guidance jumped directly to using firecrawl_agent.

Testing showed that using firecrawl_map with a search parameter can efficiently discover the specific page URL (e.g., /reference/webhook-events instead of just /reference), allowing a subsequent scrape to succeed without the overhead of the full agent.

Example workflow that now works:

1. Scrape https://docs.example.com/reference for "Invoice Deleted webhook parameters" → empty result
2. Map with {"url": "https://docs.example.com/reference", "search": "Invoice Deleted webhook"} → finds /reference/webhook-events
3. Scrape the specific URL → succeeds

Test plan

• Verify the updated tool description renders correctly in MCP clients
• Test with a JS-heavy documentation site to confirm the map→scrape workflow guidance helps agents succeed