Seekit Live Search for Openclaw

A zero-config search skill that provides AI agents with real-time web, social, and video data via a simple CLI or Python API.

lexiforest
v0.1.0
Mar 18, 2026
1
853
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install seekit-search

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install seekit-search using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).

Prefer to download?

Get the raw skill files in a ZIP archive.

What is Seekit Live Search?

Seekit Live Search is a powerful search utility designed for AI agents that require up-to-the-minute information from the internet without the friction of API keys or complex authentication. As a versatile entry in the Openclaw Skills library, it allows agents to bridge the gap between their training data and the live web, pulling context from search engines, social platforms, and video hosting sites.

By prioritizing structured JSON output, Seekit ensures that data is easily consumable by LLMs for reasoning, fact-checking, or content generation tasks. It acts as a lightweight proxy to the world's information, making Openclaw Skills more capable of handling time-sensitive queries.

Seekit Live Search Use Cases

  • Performing real-time fact-checking during autonomous research tasks.
  • Monitoring social media trends across platforms like Reddit, Threads, or Weibo.
  • Discovering video content and tutorials on YouTube and Bilibili for multi-modal analysis.
  • Gathering technical documentation or programming tutorials from diverse web providers.

How Seekit Live Search Works

  1. The AI agent identifies a need for external information not present in its local context.
  2. The agent selects an appropriate provider (e.g., Bing for general web, YouTube for video) based on the query intent.
  3. The agent executes a seekit command via the CLI or calls the Python API to fetch live results.
  4. Seekit scrapes and parses the network response into a structured format.
  5. The agent consumes the resulting JSON or Markdown to refine its response or take further action.

Seekit Live Search Setup

To integrate this search capability into your Openclaw Skills workflow, install the package via pip:

pip install seekit

No further configuration or environment variables are required to begin searching.

Seekit Live Search Data Schema & Taxonomy

Seekit returns a list of SerpItem objects. The data is highly structured to facilitate easy parsing by AI models:

Field Description Type
provider The search engine used (e.g., bing, google, reddit) String
title The headline or title of the search result String
excerpt A snippet or summary of the page content String
url The direct link to the source String
author The creator of the content (if available) String (Optional)
time Publication date of the result String (ISO 8601)

Seekit Live Search Advanced Features

  • Multi-platform support covering web, video, and social media ecosystems.
  • Formatted output options including JSON for programmatic use and Markdown for direct agent reading.
  • Provider redundancy: easily switch between engines like Brave, DuckDuckGo, or Sogou if a primary source is throttled or noisy.
  • Limit controls to manage context window usage by restricting the number of returned items.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*