What problem does it solve? Conducting thorough web research manually requires dozens of searches, source tracking, and report structuring. This Skill automates the entire pipeline: it reads a research brief file, executes web searches at a configurable depth, and saves a structured markdown report with citations. ## Core Features & Use Cases - Brief-Driven Research: Reads query, scope, context, and output preferences from a research_brief.md file in the working directory. - Four Depth Modes: Supports quick (5-8 searches), standard (10-15), deep (20-30 with a background agent), and ultradeep (40+ with two background agents) research passes. - Structured Reports with Citations: Produces a markdown report with table of contents, content-driven sections, a references list of actually visited URLs, and a mandatory meta section. - Use Case: Write a brief asking to compare two vector databases with pricing and limitations in scope, run the skill in deep mode, and receive a dated report saved to docs/research/ with all source URLs. ## Quick Start Create a research_brief.md with your query and mode, then ask the AI to run the research-brief skill to generate and save the report.
