mcpField
B
byted-byteplus-infoquest-search

byted-byteplus-infoquest-search

github.com/bytedanceProductivity47743 installsUpdated today
Repository

About byted-byteplus-infoquest-search

A Python skill for searching the web and images and extracting page content through the BytePlus InfoQuest API. It is intended for research workflows that need date or site filters and image size selection.

Key features

  1. 01Web search with date filters
  2. 02Search within a specified site
  3. 03Image search with size filters
  4. 04Extract content from a URL

Use cases

  1. 01Find recent coverage of a topic
  2. 02Search a particular site for projects
  3. 03Locate images by topic and size

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/bytedance/agentkit-samples --skill byted-byteplus-infoquest-search

After installing, type/byted-byteplus-infoquest-searchin Claude Code to confirm it loaded.

Bundled files

References and scripts installed along with SKILL.md.

LICENSE
scripts/
extract.py
search.py

Details

From the project README

Byted InfoQuest

AI-optimized web search, image search and content extraction using BytePlus InfoQuest API. Returns concise, relevant results with time filtering, site-specific search, and image search capabilities.

Tip: InfoQuest is currently not available in Mainland China regions.

Environment Variables

Before using this skill, ensure the following environment variable are set:

  • INFOQUEST_API_KEY: API key for the web search, image search and content extraction service

Web Search

python3 scripts/search.py "query"
python3 scripts/search.py "query" -d 7
python3 scripts/search.py "query" -s github.com

Image Search

python3 scripts/search.py "query" -i
python3 scripts/search.py "query" -i -z l
python3 scripts/search.py "query" -i -s unsplash.com -d 30

Options

  • -d, --days <number>: Search within last N days (default: all time)
  • -s, --site <domain>: Search within specific site (e.g., github.com, unsplash.com)
  • -i, --image: Perform image search (default: web search)
  • -z, --image-size <size>: Image size filter: l (large), m (medium), i (icon)

Extract content from URL

python3 scripts/extract.py "https://example.com/article"

Examples

Recent News Search
# Search for AI news from last 3 days
python3 scripts/search.py "artificial intelligence news" -d 3
Site-Specific Research
# Search for Python projects on GitHub
python3 scripts/search.py "Python machine learning" -s github.com
Image Search Examples
# Search for cat images
python3 scripts/search.py "cat" -i

# Search for large landscape images
python3 scripts/search.py "landscape" -i -z l

# Search for icons from specific site
python3 scripts/search.py "logo" -i -z i -s github.com

# Search for recent images
python3 scripts/search.py "sunset" -i -d 7
Content Extraction
# Extract content from a single article
python3 scripts/extract.py "https://example.com/article"

Notes

API Access
Search Features
  • Time Filtering: Use -d for searches within last N days (e.g., -d 7)
  • Site Filtering: Use -s for site-specific searches (e.g., -s github.com)
  • Image Search: Use -i for image search, with optional size filtering (-z)
  • Image Size Options: l (large), m (medium), i (icon)
Image Search Usage

Image search returns URLs to images that can be used as reference for image generation. Each result includes:

  • title: Image title or description
  • image_url: Direct URL to the image

Usage hint: Use the image_url values as reference images in image generation. Download them first if needed.

Quick Setup

  1. Set API key:

    export INFOQUEST_API_KEY=<REPLACE_WITH_YOUR_KEY>

  2. Install required Python packages:

    pip install requests

  3. Test the setup:

    # Test web search
python3 scripts/search.py "test search"

# Test image search
python3 scripts/search.py "test" -i

Error Handling

The API returns error messages starting with "Error:" for:

  • Authentication failures
  • Network timeouts
  • Empty responses
  • Invalid response formats
  • Invalid image size parameters (must be 'l', 'm', or 'i')

Differences from Node.js Version

  • Python 3.6+ required
  • requests library used instead of fetch
  • Simplified argument parsing using argparse
  • Same functionality and API endpoints

Comments

Markdown supported