jevcrawl vs ScraperAPI

jevcrawl

Three compact, typed MCP tools return verbatim passages with source URLs, cost and stop reasons, and errors arrive as codes with instructions. The project is three days old and alpha, it depends on TypeSafe's paid Jev API, which is in early access with no free tier found, and it does not block crawls of private or local addresses.

ScraperAPI

One GET request with a key and a URL returns a page, and the MCP server's 28 tools carry read-only and destructive hints. No status page, OpenAPI file, security.txt or certification was found, and the documented default puts the API key in the URL query string.

jevcrawl

1. Phrase goal as one specific question and start with max_pages between 10 and 30
2. Check stats.stopped_because and stats.degraded, and read uncertain before reporting that something is missing
3. Use scrape without a goal when you need a whole page, since it makes no paid Jev call and needs no key
4. Handle {"error": code, "message": ...} results, for example auth_error means TYPESAFE_API_KEY is missing or rejected
5. Pages in stats.needs_js_urls were not read. Tell the user rather than guessing their content

ScraperAPI

1. Set the client timeout to 70 seconds. The API retries for that long before it returns 500, and a request cancelled earlier is still charged
2. Send the key as the x-sapi-api_key header to keep it out of URLs and logs
3. Pass max_cost on each request, or call /account/urlcost first, because domain and anti-bot surcharges vary from 1 to 75 credits
4. Use output_format=markdown or text to cut response size, and autoparse=true only on supported domains
5. On 429 reduce concurrency to the plan limit (5 on Free, 20 on Hobby). The request is rejected, not queued