Firecrawl transition guide

[VinciGit00]

No description provided.

[VinciGit00]

Verification — original samples were broken on scrapegraph-py>=2.1.0

Every Python snippet in this guide used the typed-request wrappers (ScrapeRequest, ExtractRequest, SearchRequest, CrawlRequest, MonitorCreateRequest). Those classes are still importable on 2.1.0, but the SDK methods no longer accept them as the first positional argument — they expect url: str (or query, prompt, …) directly. Reproduced against the live API with the SDK at the version we ship today:

scrapegraph-py: 2.1.0
[scrape ]  RAISED ValidationError: 1 validation error for ScrapeRequest
url
  URL input should be a string or URL [type=url_type, ...]
[extract]  RAISED ValidationError: 1 validation error for ExtractRequest
prompt
  Input should be a valid string [type=string_type, ...]
[search ]  RAISED ValidationError: 1 validation error for SearchRequest
query
  Input should be a valid string [type=string_type, ...]
[crawl  ]  RAISED ValidationError: 1 validation error for CrawlRequest
url
  URL input should be a string or URL [type=url_type, ...]
[monitor]  RAISED TypeError: MonitorResource.create() missing 1 required positional argument: 'interval'

Fix in 904cd8d

• Rewrote every Python snippet to the 2.1.0 direct positional/keyword style (sgai.scrape("https://…", formats=[…]), sgai.extract(prompt, url=…), sgai.monitor.create(url, "*/30 * * * *", …), etc.).
• Bumped the SDK matrix to ≥ 2.1.0 (Python ≥ 3.12, Node ≥ 22) — was ≥ 2.0.1.
• Fixed res.data.json → res.data.json_data (the snake_case alias on ExtractResponse).
• Added tests/python-v2.1.0/test_firecrawl_transition.py that exercises every rewritten sample.

Proof — rewritten samples against the live API

[scrape ] status=success elapsed_ms=50
[extract] status=success elapsed_ms=336
[search ] status=success elapsed_ms=5498
[crawl  ] status=success elapsed_ms=152
           crawl.get -> status=success job=completed 1/2
[monitor] status=success elapsed_ms=147
           monitor.activity -> status=success
           cleaned up cron_id=1c72e4f2-1b11-4c28-95ed-48281cee50c0

All five endpoints (scrape, extract, search, crawl + crawl.get, monitor + monitor.activity + cleanup) now succeed end-to-end. Run yourself with:

SGAI_API_KEY=sgai-... ./venv/bin/python tests/python-v2.1.0/test_firecrawl_transition.py

🤖 Generated with Claude Code