Realtime Web Search for Openclaw

A high-performance search skill for AI agents that provides live web retrieval and fact-checking via enhanced Baidu routes with traceable outputs.

wangziiiiii
v1.0.1
Mar 18, 2026
1
971
0

Install & Download

1. ClawHub CLI

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

npx clawhub@latest install realtime-web-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 realtime-web-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 Realtime Web Search?

This skill enables AI agents to perform live web searches using a robust Baidu-based fallback chain. It is designed for scenarios requiring speed, cross-source verification, and an audit trail for data provenance. By integrating Openclaw Skills like this one, developers can ensure their agents access current news and changing facts that are often missing from static training data.

Unlike standard search integrations, it offers multiple routing modes including search, summary, and auto. This allows for fine-tuning between quick evidence gathering and comprehensive information synthesis, making it an essential component for research-oriented AI workflows.

Realtime Web Search Use Cases

  • Verifying rapidly changing facts or current events across multiple web sources.
  • Collecting Chinese web evidence for technical documentation, reports, or market research.
  • Fact-checking AI-generated content against live search results to reduce hallucinations.
  • Switching between high-speed search and deep summarization for complex information gathering.

How Realtime Web Search Works

  1. The agent receives a user query and triggers the Realtime Web Search skill via Openclaw Skills.
  2. The system initiates the primary web_search route for fast, structured retrieval of web results.
  3. If the primary route fails or requires more context, it falls back to a chat/completions search implementation.
  4. For tasks requiring synthesis, the web_summary route compresses information into a long-form summary.
  5. Results are returned in a structured JSON format, including a request_id and source endpoint for full traceability.

Realtime Web Search Setup

Ensure Python 3 is installed and your environment variable is set. Obtain a Bearer Token or API Key from the Baidu Intelligent Cloud console.

# Set your API Key
export BAIDU_API_KEY='your_bearer_token_here'

# Run a test query from the skill directory
py -3.11 scripts/search.py '{"query":"OpenClaw latest version","mode":"search"}'

Realtime Web Search Data Schema & Taxonomy

The skill returns a JSON array of objects with the following structure:

Field Description
source The route hit (e.g., chat, web_search, or web_summary)
source_endpoint The specific API endpoint used for the request
request_id Unique identifier for debugging and auditing
answer The synthesized response or a list of citations (title, URL, content)

Realtime Web Search Advanced Features

  • Multi-route fallback architecture ensures high availability and reliability for mission-critical tasks.
  • Advanced filtering support including site-specific matching and recency filters such as week, month, or year.
  • Website blacklisting capabilities to filter out unwanted or low-quality sources from the results.
  • Flexible mode selection (auto, search, summary) to optimize for either response speed or depth of information.
  • Full compatibility with Openclaw Skills framework for seamless AI agent integration.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Requires
Github Stars: 0
forks: 0

Featured*