Research Skill

Recursive web research and fact-checking is performed by iteratively generating diverse queries, browsing high-value links, extracting content, and cross-validating contradictions until information gaps are closed. The process prioritizes authoritative sources and escalates to the lynx text browser when deep extraction is required.

The best way to synthesize authoritative information from fragmented web sources is to use a depth-first research workflow that selectively escalates between search APIs and the lynx text browser. This approach extracts, cross-checks, and prioritizes recent sources to resolve contradictions and build a structured synthesis.

The lynx text browser works for multi-step web research queries by enabling manual, depth-first exploration of web links identified during initial search API lookups. It extracts raw text content from prioritized sources, allowing the research loop to detect gaps and validate facts directly from the target pages.

Yes, you can use this approach for technical comparisons and investigative literature reviews. The iterative research loop is designed to handle complex, fragmented, or time-sensitive questions by continuously browsing, extracting, and cross-validating data across multiple sources until the literature gaps are closed.

To resolve contradictions and validate sources during iterative web browsing, the workflow cross-checks conflicting information across multiple high-value links. It prioritizes authoritative and recent sources, using selective escalation to the lynx text browser for deeper extraction and gap detection until termination criteria are met.