# @agntn/web ## Overview - [@agntn/web](https://web.agntn.dev/raw/index.md) ## Documentation Sets - [@agntn/web](https://web.agntn.dev/llms-full.txt) ## Guide - [Getting Started](https://web.agntn.dev/raw/guide.md): Install the package, run one search, and get the same result shape from thirteen providers. - [Agents](https://web.agntn.dev/raw/guide/agents.md): The same tools for the Vercel AI SDK, the MCP server, Pi and OMP, and what an agent should and should not do with the answer. - [CLI](https://web.agntn.dev/raw/guide/cli.md): The web command, its subcommands, the flags, and the JSON envelopes that match the library. - [Custom Providers](https://web.agntn.dev/raw/guide/custom.md): Extend the base class, implement the capabilities you have, register it, and every helper, tool and CLI picks it up. - [Explorer](https://web.agntn.dev/raw/guide/explorer.md): The explorer page behind web.agntn.dev that runs the library live, and what it does to be a good citizen. - [Fan-out](https://web.agntn.dev/raw/guide/fanout.md): Every configured provider at once with deduplicated results and evidence, automatic fallback for one query, and batches of queries. - [Reverse Image Search](https://web.agntn.dev/raw/guide/image.md): A public image URL in, the pages that contain or resemble it out, through SerpAPI Google Lens. - [Reading](https://web.agntn.dev/raw/guide/read.md): URL in, normalized page out, with Jina Reader first, automatic fallback, an exact output bound and a continuation for the rest. - [Reveal](https://web.agntn.dev/raw/guide/reveal.md): The parts of a page its reader never sees. Comments, hidden form fields, invisible characters and white text on white, each with its line. - [Searching](https://web.agntn.dev/raw/guide/search.md): One provider, one query, the options every adapter takes, the filters each one honours, and the diagnostics that come back. ## Providers - [Providers](https://web.agntn.dev/raw/providers.md): Thirteen providers, one class shape. Pick one to see its endpoints, the filters it honours, the fields it fills and the traps. - [Brave](https://web.agntn.dev/raw/providers/brave.md): The Brave Search API, one GET per page, with extra snippets joined into text and a continuation that carries the offset. - [Context.dev](https://web.agntn.dev/raw/providers/context.md): The Context.dev API for search with Markdown ranked by relevance and for scraping pages into Markdown or HTML. - [Exa](https://web.agntn.dev/raw/providers/exa.md): The Exa neural search API, the richest result shape of the built-ins, with highlights by default and summaries and full text on request, and Exa Contents for reading pages. - [Firecrawl](https://web.agntn.dev/raw/providers/firecrawl.md): The Firecrawl v2 API for search with scraped passages and for scraping a page into Markdown or HTML. - [Jina](https://web.agntn.dev/raw/providers/jina.md): Jina's s.jina.ai for search and r.jina.ai for reading, the default reader, with an optional key. - [Marginalia](https://web.agntn.dev/raw/providers/marginalia.md): Marginalia Search, an independent index of small personal sites and plain text pages, usable without a key when you name it. - [Mojeek](https://web.agntn.dev/raw/providers/mojeek.md): The Mojeek Search API, an independent index with classic SERP metadata, domain and date filters and paging. - [OpenAI Codex](https://web.agntn.dev/raw/providers/openai-codex.md): Web search through your existing Pi, OMP, Codex or OpenCode login. - [SearXNG](https://web.agntn.dev/raw/providers/searxng.md): A self-hosted SearXNG metasearch instance, no key, with a reachability probe and the engines behind each result in metadata. - [SerpAPI](https://web.agntn.dev/raw/providers/serpapi.md): SerpAPI's Google endpoints for classic SERP results and Google Lens for reverse image search. - [SerpBase](https://web.agntn.dev/raw/providers/serpbase.md): SerpBase's Google SERP endpoints, with a category that picks the web, images, news or videos endpoint. - [Tavily](https://web.agntn.dev/raw/providers/tavily.md): The Tavily search API, with an optional generated answer in response metadata, and Tavily Extract for reading pages. - [TinyFish](https://web.agntn.dev/raw/providers/tinyfish.md): TinyFish Search for news and research results with publisher metadata, and TinyFish Fetch for reading pages. ## Tools - [Explorer](https://web.agntn.dev/explorer): Search one provider, fan out to every configured one, or read a page through the docs worker and see the normalized answer.