Synchronous client for the wigolo REST API. Args: base_url: Server base URL. Falls back to ``WIGOLO_BASE_URL`` then ``http://

Async MCP client that communicates with wigolo via subprocess stdio. Spawns `npx wigolo` as a child process, sends JSON-RPC 2.0 requests over

Async MCP client that communicates with wigolo via subprocess stdio. Spawns `npx wigolo` as a child process, sends JSON-RPC 2.0 requests over

Read web pages via wigolo fetch and convert to LlamaIndex Documents. Each URL is fetched through wigolo's MCP server. The extracted markdown

Retriever that uses wigolo's search tool to find relevant documents. Each search result is converted to a LangChain Document with the page's

Fetch a web page using wigolo and return the result as JSON. Wraps wigolo's fetch MCP tool. Returns a JSON string with the page title, clean

Search the web via wigolo and convert results to LlamaIndex Documents. Each search result becomes a Document with its markdown content (or snippe

Search the web using wigolo and return results as JSON. Wraps wigolo's search MCP tool. Returns a JSON string with search results including t

Base class for all errors raised by this SDK.

Raised when the MCP client encounters an error.

Raised when the MCP client encounters an error.

Async client for the wigolo REST API. Wraps a synchronous ``Client`` and dispatches each blocking request to a bounded ``ThreadPoolExecutor``

Handle to a local wigolo daemon (spawned or reused).

A non-2xx HTTP response from the wigolo REST API. ``error``/``error_reason``/``stage`` come from the server's error envelope when present. ``

A transport-level failure that produced no HTTP response. Most commonly a refused connection. For zero-setup local use, route through ``wigol

Crawl a site from a seed URL.

Extract structured data from a page.

Fetch a specific web page as clean markdown.

Run multi-step research and return a structured brief.

Search the web for information on any topic.