Search every provider at once - Google, Anthropic, OpenAI, Perplexity, X.AI and Brave.

npm install @just-every/search

import { web_search } from '@just-every/search';

// Basic usage (backward compatible)
const results = await web_search('brave', 'your search query', 5);

// With agent tracking (for MAGI system)
const results = await web_search('agent-id', 'brave', 'your search query', 5);

# Install globally
npm install -g @just-every/search

# Search with Brave (default)
web-search "your search query"

# Use a specific engine
web-search "your search query" -e anthropic

# Get more results
web-search "your search query" -n 10

# Output raw JSON
web-search "your search query" --json

# Run comprehensive research with AI agents
web-search task "your research query"

# Use a specific model class for research
web-search task "your research query" -m reasoning

# Available model classes: standard, mini, reasoning, reasoning_mini (default),
# monologue, metacognition, code, writing, summary, vision, vision_mini,
# search, image_generation, embedding, voice

• brave - Privacy-first search using Brave's independent index
• anthropic - Deep multi-hop research with strong source citations (requires ANTHROPIC_API_KEY)
• openai - ChatGPT-grade contextual search (requires OPENAI_API_KEY)
• google - Fresh breaking-news facts via Gemini grounding (requires GOOGLE_API_KEY)
• sonar - Lightweight Perplexity search (requires OPENROUTER_API_KEY)
• sonar-pro - Advanced Perplexity search (requires OPENROUTER_API_KEY)
• sonar-deep-research - Expert-level Perplexity research (requires OPENROUTER_API_KEY)
• xai - Real-time web search via Grok (requires XAI_API_KEY)

Set the following environment variables for the search engines you want to use:

• BRAVE_API_KEY - For Brave search
• ANTHROPIC_API_KEY - For Anthropic search
• OPENAI_API_KEY - For OpenAI search
• GOOGLE_API_KEY - For Google search
• OPENROUTER_API_KEY - For Perplexity searches
• XAI_API_KEY - For X.AI search

Performs a web search using the specified engine.

• inject_agent_id (optional) - Agent ID for tracking in MAGI system
• engine - Search engine to use
• query - Search query string
• numResults - Number of results to return (default: 5)

Returns a JSON string containing search results.

Runs comprehensive research using multiple search engines in parallel with AI agents.

• query - Research query string
• modelClass - Model class to use (default: 'reasoning_mini')

Returns a comprehensive research report as a string.

Returns an array of ToolFunction definitions for use with the MAGI agent system.