baidu-search

What This Skill Does

The baidu-search skill serves as a powerful bridge between the OpenClaw AI agent and the vast index of the Baidu AI Search Engine (BDSE). By leveraging this tool, your AI agent gains the ability to retrieve real-time data, verify facts against current events, and perform complex research tasks that require up-to-date web connectivity. Unlike basic search utilities, this skill offers granular control over result sets, allowing users to define specific resource types—such as prioritizing video content over textual articles—or filtering by temporal relevance using advanced date range parameters. It is an essential component for agents tasked with maintaining context on evolving topics or localized information relevant to the Chinese web ecosystem.

Installation

To integrate the Baidu Search capability into your OpenClaw environment, execute the following command in your terminal:

clawhub install openclaw/skills/skills/ide-rea/baidu-search

Ensure that you have the necessary environment permissions for network egress, as the skill requires active communication with Baidu's API endpoints to function correctly.

Use Cases

This skill is ideal for research-heavy AI workflows. Use it for:

• Monitoring breaking news or trending topics within the Chinese internet space.
• Aggregating specific documentation or knowledge base articles from curated domains.
• Performing multi-modal data gathering, such as pulling both web text and related video content for a single subject.
• Filtering historical data through site-specific or time-bound queries, which is vital for academic or competitive intelligence tasks.

Example Prompts

1. "Perform a search for the latest developments in generative AI released in the last month, limiting results to news.baidu.com."
2. "Find me the top 10 video tutorials regarding Python programming basics using the baidu-search skill."
3. "Research current environmental policies in Beijing and provide a summary based on the web results."

Tips & Limitations

To get the best results, use the search_recency_filter to narrow down content to relevant timeframes, as broad queries may yield outdated information. When searching for specific company data or documentation, utilize the search_filter with the match.site parameter to drastically improve signal-to-noise ratios. Note that block_websites is a highly effective way to sanitize results if you find your agent is frequently pulling content from low-quality or irrelevant sources like Tieba. Keep in mind that resource limits (e.g., max 50 for web) exist to ensure fast response times from the API; always match your top_k settings to your actual data needs to prevent unnecessary latency.