Docs
English
/Core Features/Browser Search

Browser Search

Browser Search is a zero-configuration deep search capability built into the OpenClaw Gateway. It requires no API Key and automatically performs web search, page parsing, content extraction, and result return via the built-in Chromium browser, providing stable, high-quality real-time information access for users both inside and outside mainland China.

Core Functionality Flow & Experience

When a user asks a question requiring web lookup, the system automatically executes the full workflow:

  • Task Understanding & Search Planning: The Agent automatically parses requirements and generates multi-angle search keywords.
  • Automatic Search Engine Selection: Intelligently chooses engines based on mainland China/overseas environments without manual switching.
  • Ad Filtering & Content Extraction: Automatically removes ads, accesses pages, and extracts clean main content.
  • Structured Result Return: Returns title, URL, snippet, and content to support direct report generation.
  • Full Process Unattended: One-stop completion from search to summary with real-time progress display.

Core Capability Modules

  • No API Key Required: Out-of-the-box, independent of third-party keys.
  • Automatic Environment Detection: Identifies mainland China/overseas mode based on time zone, locale, and network status.
  • Automatic Engine Fallback: Switches to the next available engine if the current one returns no results or fails.
  • Mainland China Engine Chain: Bing China → 360 Search → Baidu → Sogou.
  • Overseas Engine Chain: Google → Bing International → DuckDuckGo.

2. Deep Content Extraction

  • Automatic Ad Link Filtering: Identifies and skips advertising, promotion, and tracking links.
  • Anti-Crawl & Error Page Detection: Automatically skips login walls, 403, 404, 5xx, and other abnormal pages.
  • Smart Body Extraction: Filters out navigation, footers, ads, and retains core content.
  • Fallback on Extraction Failure: Falls back to search snippets if crawling fails, ensuring usability.
  • LRU Cache Mechanism: Returns cached results immediately for repeated queries within 5 minutes.
  • Human Behavior Simulation: Adds proper access delays to reduce blocking risks.
  • Concurrent Page Extraction: Supports simultaneous crawling of multiple pages for efficient deep search.
  • Result Truncation Control: Automatically limits per-page content length for stable, efficient responses.

4. Standardized Results & Compatibility

  • Uniform Output Format: Returns standardized structure with title, url, snippet, and content.
  • Compatible with Existing Plugins: Works directly with Agents without modification.
  • Baidu Redirect Handling: Automatically resolves encrypted links to obtain real target URLs.
  • User-Friendly Error Messages: Returns clear prompts when engines are unavailable for easy troubleshooting.

Typical Use Cases

  • Enterprise / Industry Information Research: Automatically investigates business, products, financial reports, and competitor moves.
  • Real-Time Information Access: News, policies, event progress, and impact analysis.
  • Knowledge & Data Supplement: Retrieves latest documents, data, reports, and technical materials.
  • Mainland China No‑API‑Key Scenarios: Provides stable web search for users without API keys.
  • Automated Report Generation: Summarizes multi-round searches and outputs structured analysis directly.

Core Value Summary

SingClaw Browser Search delivers zero-configuration, cross-environment, fully automated deep search capabilities. It solves pain points including cross-region network limitations, cumbersome API Key setup, cluttered search results, and difficult content extraction. It equips Agents with stable, high-quality real-time information access, truly achieving out-of-the-box, one-click deep search.

Screenshot & Analysis
SingMemory