Pi Web Access
A web search, content extraction and video understanding extension for the Pi coding agent, with zero-config search and fallback chains across more than twenty search providers.
Pi Web Access is an extension for the Pi coding agent that adds web search, page content extraction, GitHub repository cloning, and video understanding. It works out of the box with no API key needed through a zero-config Exa search, and can reuse an existing Codex or Kimi Code Plan subscription's authentication for their respective search backends.
Additional providers, including OpenAI, Brave, Parallel, Tavily, Firecrawl, Jina, Kagi, SerpApi, Perplexity, Gemini and Mistral, can be added with API keys or endpoints for more control, and a self-hosted SearXNG or Crawl4AI endpoint can be configured for private search and extraction. Each capability has a fallback chain, so search, page fetching and YouTube understanding all try several providers in a defined order until one succeeds; GitHub URLs are cloned locally instead of being scraped, giving the agent real file contents.
It installs with pi install npm:pi-web-access and is released under the MIT license.