I used Serp via API many moons ago. The most interesting part of the company imo is their legal defense of different plans:
Production - $150
15,000 searches / month
U.S. Legal Shield
ie. "Our U.S. Legal Shield protects your right to crawl and parse public search engine data under the First Amendment. We assume scraping and parsing liability for customers on most recurring plans unless your usage is illegal."
I imagine at least some portion of companies use them just for this liability shield.
Sounds a lot like the old guarantee paid SSL certificate providers used to offer; pretty words, but meaningless in practice. (IIRC, no one ever got a payout from any of them.)
"We assume scraping and parsing liabilities for both domestic and foreign companies unless your usage is otherwise illegal" seems like a big loophole in it.
Couldn't this be laid out as, We assume scraping and parsing liability unless it is ruled as being illegal, in which case your use would be illegal and our liability shield wouldn't help you?
Hi, thanks for the feedback. We actually already support `Bearer: Authorization <token>` auth scheme. It is accessible on https://mcp.serpapi.com/mcp path.
I would say the /apiKey/mcp integration path is more tailored for manual Claude Desktop / VS Code users where it might be easier, than trying to insert headers somehow.
Regarding the environment variables, the MCP is stateless so the secret management is handled on the client and obviously it could be set via env vars.
I used Serp via API many moons ago. The most interesting part of the company imo is their legal defense of different plans:
ie. "Our U.S. Legal Shield protects your right to crawl and parse public search engine data under the First Amendment. We assume scraping and parsing liability for customers on most recurring plans unless your usage is illegal."I imagine at least some portion of companies use them just for this liability shield.
Sounds a lot like the old guarantee paid SSL certificate providers used to offer; pretty words, but meaningless in practice. (IIRC, no one ever got a payout from any of them.)
"We assume scraping and parsing liabilities for both domestic and foreign companies unless your usage is otherwise illegal" seems like a big loophole in it.
Couldn't this be laid out as, We assume scraping and parsing liability unless it is ruled as being illegal, in which case your use would be illegal and our liability shield wouldn't help you?
> unless your usage is illegal
Like copyright infringement of Google's search results?
Embedding the API key (secret) in the URL complicates secret management. Please move that to a header and allow setting via env var.
Hi, thanks for the feedback. We actually already support `Bearer: Authorization <token>` auth scheme. It is accessible on https://mcp.serpapi.com/mcp path.
I would say the /apiKey/mcp integration path is more tailored for manual Claude Desktop / VS Code users where it might be easier, than trying to insert headers somehow.
Regarding the environment variables, the MCP is stateless so the secret management is handled on the client and obviously it could be set via env vars.