Find the passages of an article that answer a question

Read a public article or YouTube URL and return only the passages relevant to your query, each anchored to its paragraph and labelled with the section it sits u

liveunknown~5s typical

mcpsmryproduct

$0.0020

per call

plus $0.001 platform fee per run

Open MCP server; Cracked bills routing only

Provider Upstream docs

$Set up https://cracked-ai-pearl.vercel.app/SKILL.md, then use Cracked to run mcp-ai-smry-r-smry-product /search_article for me.

Input

Field	Type	Description
url*	string	The public http or https article or YouTube URL to read.
query*	string	What you want to find in the article. Natural language or keywords; matching is lexical, s
max_tokens	integer	Approximate token budget for the returned text. Defaults to 4000, which covers a typical a
max_passages	integer	Maximum passages to return. Defaults to 8.
cursor	integer	Offset returned as next_cursor by an earlier search.

Call it

curl

curl https://cracked-ai-pearl.vercel.app/v1/run \
  -H "Authorization: Bearer ck_live_..." -H "content-type: application/json" \
  -d '{"provider":"mcp-ai-smry-r-smry-product","endpoint":"/search_article","input":{"url":"...","query":"..."}}'

cli

npx cracked-ai run -p mcp-ai-smry-r-smry-product -e /search_article -i '{"url":"...","query":"..."}'

mcp

run_tool({ provider: "mcp-ai-smry-r-smry-product", endpoint: "/search_article", input: {"url":"...","query":"..."} })

Try it

Runs against your signed-in workspace balance. Sign in if you have not.