Wigolo Ships Local Web Search and Crawling for AI Agents

Your AI agent's web search has a meter running on every query. Wigolo replaces that with a local search stack that needs no API key and never charges per query.

What Happened

Wigolo fans out queries across 18 search engines at once, fuses the results with ML reranking, and caches everything so re-querying costs nothing. Each result carries a verbatim excerpt pinned to its exact byte position in the source, an explainable score decomposition, and live engine telemetry.

A single install ships eight tools: Search, Fetch, Crawl, Extract, Cache, Find Similar, Research, and Agent. Search hits 18 direct engine adapters, Fetch auto-escalates to a headless browser on anti-bot challenges, and Crawl supports BFS, DFS, sitemap, or map-only modes with per-domain rate limits and robots.txt compliance.

It runs locally and connects to Claude Code, Cursor, Codex, Gemini CLI, VS Code, Windsurf, and Zed through MCP. There is also a REST API, TypeScript and Python SDKs, and framework integrations for LangChain, CrewAI, LlamaIndex, and the Vercel AI SDK. The project is available on GitHub.

Why It Matters

Firecrawl, Exa, and Tavily all require an API key and charge per query, and agents that search in bursts compound that cost fast. Wigolo removes the meter and keeps queries on your machine instead of sending data out on every call. If your agent does research, crawling, or extraction, that is both a cost fix and a privacy win in one install.