Wigolo - an open-source local search tool that supports integration with various programming agents.
Wigolo is an open-source local search tool that integrates with programming agents such as Claude Code and Cursor via MCP. It requires no API key and allows for local searching, web scraping, full-site crawling, structured data extraction, and more, all at zero cost.
What is Wigolo?
Wigolo is an open-source local search tool that integrates with programming agents such as Claude Code and Cursor via MCP. It requires no API key and allows for zero-cost local searching, web scraping, full-site crawling, structured data extraction, caching, and research. The tool supports parallel queries across 18 search engines, and results include interpretable scores and original text citations; all data is stored locally.
Wigolo's main functions
-
searchIt calls 18 search engines in parallel and returns search results with interpretable scores through on-device ML reordering.
-
fetchLayered routing for web page crawling, automatically upgrading from pure HTTP to a headless browser, intelligently dealing with anti-crawling measures and SPA pages.
-
crawlSupports full-site crawling using BFS, DFS, and sitemap modes, adheres to robots.txt, and automatically removes duplicates.
-
extractExtract structured data from the page, including tables, metadata, JSON-LD, and custom JSON schemas.
-
cacheIt stores the crawled content in a local index, supports keyword and semantic search, provides millisecond-level response for repeated queries, and is available offline.
-
find_similarBy integrating keywords, semantics, and real-time webpage data, it finds pages similar to the target URL or concept.
-
researchIt automatically decomposes complex problems into subqueries, acquires resources in parallel, and then synthesizes a structured research report with cited sources.
-
agentIt executes an autonomous collection cycle of "plan → search → acquire → extract → synthesize", with a step log and time budget.
-
diff + watchMonitor changes to the page since the last visit and push the changes to the specified webhook.
Follow us on WeChat and reply with "open source",join inAI open source project discussion group
How to use Wigolo
-
Environmental preparationEnsure that Node.js version 20 or above is installed locally.
-
One-click installation:run
npx wigolo initAutomatically downloads browser engines and models, and integrates with agents such as Claude Code and Cursor. -
Health check:implement
npx wigolo healthVerify that the search, browser, and device models are functioning correctly. -
Optional LLM configurationFor research or agent functionality, set up a free Gemini API Key or a local Ollama model.
-
Calling in AgentIn agents that support MCP, such as Claude Code and Cursor, you can directly invoke search and crawling tools using natural language.
-
CLI direct useRun via command line
npx wigolo search/fetch/crawlThe specific operation is executed according to the instructions. -
Local cache reuseAll data is automatically stored
~/.wigolo/The directory supports offline reuse and semantic search.
Wigolo's core advantages
-
Local priorityAll searching, fetching, and caching are done locally, data does not leave the device, and it can be used normally even when the network is disconnected.
-
Zero cost and zero thresholdNo account registration or API key required; every query is permanently free; and a single command is all you need to connect to mainstream agents.
-
The results are fully explainable.Each search result includes a byte-level excerpt of the original text, citation ID, and rating breakdown, making the quality level immediately apparent.
-
Multi-engine parallel searchSimultaneous queries from 18 search engines, combined with ranking fusion and on-device machine learning reordering, ensure comprehensive and accurate results.
-
Intelligent hierarchical crawlingAutomatic upgrades from pure HTTP to TLS spoofing and headless browsers, honestly combating anti-scraping measures and clearly notifying users of any failures.
-
persistent local cacheAll content is stored in a local hybrid index (keywords + semantic vectors), and repeated queries receive millisecond-level responses.
Wigolo's project address
- Project official website:https://knockoutez.github.io/wigolo/
- GitHub repository:https://github.com/KnockOutEZ/wigolo
Comparison of similar products from Wigolo
| Comparison Dimensions | Wigolo | Firecrawl |
|---|---|---|
| Project Positioning | Local-first AI Agent Network Intelligence Layer | Cloud-based web crawling and structured extraction platform |
| Core propositions | Searching, crawling, and research all happen locally, at zero cost. | Convert any website into structured data usable by LLM. |
| Developers | @KnockOutEZ (Personal open source) | Mendable (a funded startup) |
| license | AGPL-3.0 (Permanently open source and free) | Closed-source commercial products |
| Search capabilities | 18 search engines running in parallel, on-device machine learning re-ranking, results with interpretable scores. | There is no built-in search function; it requires the use of other search APIs. |
| Web scraping | Layered routing (HTTP → TLS masquerading → headless browser) | Powerful fetching capabilities, supporting dynamic rendering |
| Full site crawling | BFS/DFS/sitemap, comply with robots.txt | Supports full-site crawling and site map generation |
| Structured extraction | Tables, metadata, JSON-LD, custom JSON schemas | Supports LLM extraction and structured data output |
| Local cache | Persistent local index (keywords + semantic vectors), available offline, free for repeated queries. | There is no local cache; each fetch uses the cloud API. |
| Interpretability of results | Byte-level original text excerpt + Citation ID + Rating breakdown | Returns structured data, without item-by-item scoring and source identification. |
Application scenarios of Wigolo
-
AI Programming AssistanceThe Agent automatically searches for the latest API documentation, official tutorials, and best practices while you are writing code, without requiring you to manually switch browsers.
-
Technical research and competitor analysisThe Agent autonomously searches and crawls multiple technical blogs and documentation sites to generate a structured research report with citations.
-
Full site data crawlingBatch crawl product catalogs, pricing information, academic papers, or news sites, and extract tabular and structured data to local storage.
-
Local knowledge base constructionCache frequently accessed documents and web pages locally to create an offline, searchable knowledge base for individuals or teams.
-
Page Change MonitoringContinuously monitor competitors' official websites, document update pages, or pricing pages, and automatically notify them via webhook when changes occur.