A professional web search skill that integrates Baidu Qianfan to provide real-time snippets, titles, and URLs for Chinese web content.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install jisu-baidu
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install jisu-baidu 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 Baidu Web Search?
Baidu Web Search is a robust Openclaw Skills extension powered by the Baidu Qianfan AppBuilder API. It allows AI agents to perform live web queries, returning structured information including snippets, titles, and direct links. This skill is specifically designed for scenarios where an agent needs to cite sources or provide users with a list of relevant websites for manual verification.
Unlike direct LLM generation, this skill ensures that the agent's responses are grounded in current, real-world data from the Chinese internet. It is highly effective when paired with other Openclaw Skills like JisuAPI to combine broad web knowledge with authoritative structured data such as weather, finance, or automotive specifications.
Baidu Web Search Use Cases
- Searching for the latest Chinese government policies or official corporate websites.
- Retrieving recent news and summaries from the past week, month, or year.
- Compiling a list of reference URLs and snippets for research purposes.
- Performing site-specific searches on domains like JisuAPI or government portals.
- Cross-referencing AI-generated claims with live web results for higher accuracy.
How Baidu Web Search Works
- The AI agent identifies a user intent requiring external web knowledge (e.g., "Find the latest news on...").
- The skill triggers a POST request to the Baidu Qianfan API using the provided search query and filters.
- The Baidu engine processes the request, applying parameters like recency filters or site-specific constraints.
- The skill receives a JSON response containing a list of reference objects (ID, title, URL, snippet, and date).
- The agent parses this structured data to summarize the findings or present the source list to the user.
Baidu Web Search Setup
To use this skill within the Openclaw Skills ecosystem, you must configure your Baidu AppBuilder API key as an environment variable.
# Linux / macOS
export BAIDU_API_KEY="your_appbuilder_api_key_here"
# Windows PowerShell
$env:BAIDU_API_KEY="your_appbuilder_api_key_here"
You can then run the skill via CLI:
python3 skills/baidu/baidu.py search '{"query":"AI technology trends","top_k":5}'
Baidu Web Search Data Schema & Taxonomy
The skill utilizes a structured JSON request and response format to ensure compatibility with Openclaw Skills workflows.
Request Parameters
| Field | Type | Required | Description |
|---|---|---|---|
query |
string | Yes | The search term to be queried. |
top_k |
int | No | Number of results to return (default: 20). |
sites |
array | No | Limit search to specific domains. |
search_recency_filter |
string | No | Time range: week, month, semiyear, year. |
Response Structure
references: An array containingtitle,url,snippet, anddate.request_id: A unique identifier for the API transaction.
Baidu Web Search Advanced Features
- Site-specific filtering to restrict results to trusted or niche domains.
- Recency filtering (search_recency_filter) to prioritize fresh content over historical data.
- Safe search toggling for content moderation.
- Raw resource type filtering for power users to customize the Baidu Qianfan engine behavior.
- Error handling that maps Baidu API codes to standard Openclaw Skills error structures for seamless agent recovery.
SKILL.md
Loading
Related Openclaw Skills
OCR.space Text Recognition Skill
A high-performance text recognition tool that enables AI agents to extract content from images across 20+ languages using the OCR.space API.
PromptGit Local Prompt Version Control
A local, dependency-free version control system designed to track, diff, and rollback iterations of AI prompts and templates.
CyberWin Smart Access Parameter Converter
A specialized string processing engine designed to map dynamic hardware parameters to CyberWin smart access control templates.
Hello World Skill
A foundational template designed to demonstrate the core architecture and execution flow of a custom AI agent capability.
Clawlendar
Clawlendar is a timestamp-first perpetual calendar bridge that enables AI agents to perform complex cross-calendar conversions and timeline normalization.
Quantum Circuit Simulator
A high-performance, Python-based quantum statevector simulator for exploring quantum algorithms and gate operations directly within your AI agent environment.
Featured*
