createContextFetchSearchTool() (id contextfetch-search): web search; each result carries its title, URL, published date and page content as Markdown (by default only the parts relevant to the query).
createContextFetchContentsTool() (id contextfetch-contents): reads web pages as Markdown; with a query, only the parts of each page relevant to it.
createContextFetchTools(): both, with shared options.

Install

npm install @contextfetch/mastra @mastra/core zod

Set CONTEXTFETCH_API_KEY (get a key). Without a key, requests use the keyless tier (a small daily allowance per IP).

Usage

import { Agent } from "@mastra/core/agent";
import { createContextFetchTools } from "@contextfetch/mastra";

export const researchAgent = new Agent({
  id: "research-agent",
  name: "Research Agent",
  model: "openai/gpt-5-mini",
  instructions: "Search the web, read the best sources, then answer with their URLs.",
  tools: createContextFetchTools(),
});

Options

All factories take apiKey (default: the CONTEXTFETCH_API_KEY env var), baseUrl and timeoutMs. The search tool also takes:

numResults: results per search, 1 to 25 (default 5).
content: "prunedContext" (default, the parts of each page relevant to the query), "text" (the whole page) or "highlights" (a few key sentences).

The agent can set query, includeDomains, excludeDomains and startPublishedDate on each search, and urls (up to 10) and query on each read. Both tools have Zod input and output schemas. API errors are thrown as ContextFetchError from @contextfetch/sdk.

Development

npm install
npm test                                   # unit tests against a local fake API, including a Mastra agent run
CONTEXTFETCH_API_KEY=... npm test          # plus the live smoke test (about 2 credits)