Can I search both my local project and the internet?

Yes, the skill supports both 'local' and 'web' modes. It can search your local filesystem using tools like Grep and Glob, or search the live web using WebSearch and WebFetch.

Which web search providers are supported?

It currently uses built-in WebSearch and WebFetch tools, with planned future integrations for specialized providers like Jina.ai, Firecrawl, and Exa.ai.

Does this skill require me to provide input files?

No, the Search skill is designed to be 'input-less.' You simply provide a mission description, and the skill autonomously identifies and retrieves the necessary data using its internal tools.

What format are the search results delivered in?

All findings are compiled into a structured JSON schema. This includes the source (file path or URL), content snippets, relevance scores, and metadata like line numbers or dates.

Universal Search & Data Retrieval

Name: Universal Search & Data Retrieval
Author: memorysaver

bymemorysaver

•

웹 스크래핑 및 데이터 수집

Executes autonomous search missions across local codebases and the web to return structured, attributed data.

소개

The Search skill is an autonomous, input-less capability for Claude Code designed to streamline complex information gathering. By bridging local filesystem exploration with real-time web retrieval, it allows AI agents to scan repositories, monitor trending news, or extract specific technical data without requiring manual file inputs. It intelligently manages tool selection—utilizing Glob and Grep for local files or WebSearch for online content—and compiles findings into a unified, machine-readable JSON schema complete with source attribution and relevance scoring.

주요 기능

Unified interface for both local filesystem and web-based searches
Deep codebase scanning using Glob, Grep, and Bash tools
Autonomous execution based on mission descriptions without manual file inputs
Structured JSON output including metadata, source attribution, and relevance scores
10 GitHub stars
Web content extraction and real-time URL fetching capabilities

사용 사례

Gathering documentation and technical data from external URLs for project context
Scanning a repository for specific code patterns, TODOs, or architectural implementations
Automating market research or news monitoring from sites like Hacker News

소개

주요 기능

Unified interface for both local filesystem and web-based searches
Deep codebase scanning using Glob, Grep, and Bash tools
Autonomous execution based on mission descriptions without manual file inputs
Structured JSON output including metadata, source attribution, and relevance scores
10 GitHub stars
Web content extraction and real-time URL fetching capabilities

사용 사례

Gathering documentation and technical data from external URLs for project context
Scanning a repository for specific code patterns, TODOs, or architectural implementations
Automating market research or news monitoring from sites like Hacker News