deep-research

Automate web research by breaking down complex queries into sub-questions, executing parallel web searches, analyzing coverage gaps, and synthesizing cited reports. This Skill manages the full research lifecycle from iterative query refinement to final data synthesis.

Query decomposition breaks a complex research question into multiple targeted sub-questions. This allows the system to execute parallel web searches concurrently, gathering broader data coverage before analyzing gaps and synthesizing the findings.

Yes, you need the web_search tool configured. This Skill uses Pieces MCP for web search execution, which explicitly requires the web_search dependency to gather results and facilitate the self-evolution of research strategies.

Iterative refinement analyzes initial search coverage to detect knowledge gaps, then automatically generates and executes follow-up queries. This continuous process ensures exhaustive data gathering before synthesizing the final comprehensive report.

Yes, the system synthesizes findings from parallel web searches into a comprehensive, cited report. It automatically processes the gathered data, detects coverage gaps, and compiles the synthesized information with proper citations.

Automated research synthesis depends entirely on the web_search tool's capabilities and query clarity. It may hit constraints with highly specialized niche topics or limited online data coverage, requiring manual follow-up queries to fill undetected gaps.