Agent builders
Give assistants a search layer that returns tools, docs, packages, and provider pages they can act on.
Query Here helps agents find usable websites, APIs, CLIs, MCP servers, and docs, then labels the access path and human steps needed.
Agent CLI
Search from a terminal and hand direct-use results back to your agent.
npm install -g @queryhere/cliAgent API
Give your agent a token for hybrid search, access labels, and human-step guidance.
Agent MCP
Connect Claude Desktop or Cursor instantly via the Model Context Protocol.
npx -y @queryhere/mcp-serverWho this is for
Give assistants a search layer that returns tools, docs, packages, and provider pages they can act on.
Expose reliable discovery for APIs, MCP servers, CLIs, SDKs, and browser flows without curating every result by hand.
See confidence, evidence, and human-required labels before a result is trusted in an automated workflow.
Why the problem matters
Generic search ranks pages for people who can skim, compare, and ignore noise. Agents often have to fetch those pages, spend tokens reading them, and infer whether an API, package, MCP server, or usable product flow exists. Query Here returns access labels and evidence first so agents can move toward action instead of reading another article.
Ask for a task, tool, API, CLI, MCP server, SDK, docs, or website your agent can actually use.
Results are labeled as API, CLI, MCP, SDK, docs, browser flow, login, or payment instead of plain links.
Provider and tool pages are favored over articles, reviews, and news when the agent needs to act.
Need the catalog?
The homepage is for setup. The registry is for browsing stored services and provider pages.
FAQs
No. It is a search layer for agents and developers who need direct-use destinations, access metadata, and evidence links.
It lets MCP clients call Query Here search from tools such as Claude Desktop and Cursor using an API key.
No. Verified catalog entries are labeled, and live or cached candidates keep confidence and evidence fields visible for review.