What problem does it solve? Answering broad or complex questions with a single web search yields shallow, unverified results. This Skill decomposes a topic into independent subtopics, dispatches parallel research agents, and synthesizes cross-referenced findings into a single cited report saved to disk. ## Core Features & Use Cases - Topic Decomposition with Approval Gate: Breaks a research question into 3-7 independent subtopics and requires explicit user approval before dispatching researchers. - Parallel Multi-Agent Research: Launches one pipeline-web-researcher agent per subtopic, each consulting 20+ pages across multiple search queries with source tracking and confidence ratings. - Synthesized Report: Cross-references findings, surfaces contradictions and gaps, and writes a structured report with executive summary, methodology, and deduplicated sources to .mz/research/. - Use Case: Ask "what is the state of WebAssembly in 2026" and receive a saved markdown report covering runtime performance, language support, browser adoption, and production usage with cited sources. ## Quick Start Ask the assistant to run deep research on a topic such as "research the current state of edge computing platforms" and approve the proposed decomposition when prompted.
