AgentSkillsCN

exa-synergy

借助 Exa AI SDK 进行神经网络式网络研究。系统会自动选择相应策略：快速搜索以获取即时解答，深度研究以实现全面分析，数据收割则用于收集各类清单与数据库。适用于用户需要网络搜索、文献调研、数据采集或信息挖掘的场景。

SKILL.md
--- frontmatter
name: exa-synergy
description: Neural web research with Exa AI SDK. Auto-selects strategy: fast search for quick answers, deep research for comprehensive analysis, harvest for collecting lists/databases. Use when user needs web search, research, data gathering, or information discovery.
version: 1.0.0
author: Agent Zero Custom
tags: [search, research, web, exa, ai, data-collection]
trigger_patterns:
  - "search web"
  - "research"
  - "find information"
  - "harvest data"
  - "web search"
  - "look up"
  - "find online"
allowed_tools:
  - code_execution_tool
  - memory_save
  - response

Exa Synergy — Neural Web Research

Intelligent web research tool using Exa AI SDK (v2.3+) with automatic strategy selection.

Installation

bash
# Install dependencies
pip install -r requirements.txt

# Or use setup script
bash /a0/usr/skills/setup.sh

# Set API key (get at https://exa.ai)
export EXA_API_KEY="your_api_key_here"
# Or add to /a0/.env: EXA_API_KEY=your_api_key_here

When to Use

Use this skill when you need to:

  • •Search the web for information
  • •Research a topic in depth
  • •Collect lists of companies, emails, or data
  • •Find documentation or tutorials
  • •Gather competitive intelligence

Strategy Auto-Selection

The tool automatically selects the best strategy based on query:

StrategyUse CaseExample Queries
fastQuick answers"What is...", "How do I..."
deepComprehensive research"Explain in detail...", "Research...", "Analyze..."
harvestData collection"List all...", "Find companies...", "Collect emails..."

Usage

Via Python Script

bash
python /a0/usr/skills/exa-synergy/scripts/exa_synergy.py --goal "your research goal" --mode auto --num_results 5

Parameters

ParameterTypeDefaultDescription
goalstrrequiredWhat you want to research
modestrautoStrategy: auto/fast/deep/harvest
num_resultsint5Number of results (1-20)
save_tostroptionalFile path to save results

Examples

  1. •

    Quick search:

    bash
    python /a0/usr/skills/exa-synergy/scripts/exa_synergy.py --goal "Python async best practices" --mode fast
  2. •

    Deep research:

    bash
    python /a0/usr/skills/exa-synergy/scripts/exa_synergy.py --goal "Compare React vs Vue performance" --mode deep --num_results 10
  3. •

    Data harvesting:

    bash
    python /a0/usr/skills/exa-synergy/scripts/exa_synergy.py --goal "AI startups in Europe 2024" --mode harvest --save_to /a0/tmp/research.md

Requirements

  • •EXA_API_KEY must be set in environment or /a0/.env file
  • •exa-py>=2.3.0 (installed via requirements.txt)

Output Format

Returns structured results with:

  • •Title and URL for each result
  • •Content summary
  • •Relevance score
  • •Publication date (when available)

Files

code
/a0/usr/skills/exa-synergy/
├── scripts/
│   └── exa_synergy.py
├── requirements.txt
└── SKILL.md