What problem does it solve? One-shot web searches give shallow, unverified answers. This Skill runs multi-pass research across web search, social recency sources, memory, and Notion, then synthesizes findings with citations, contradictions, and confidence levels so decisions rest on corroborated evidence. ## Core Features & Use Cases - Multi-source discovery: Combines WebSearch, WebFetch, agent-browser, /last30days (Reddit/X/YouTube/HN), local memory, and Notion, running independent passes in parallel. - Structured synthesis: Groups findings by theme, flags contradictions between sources, assigns confidence levels, and lists gaps and recommended next steps. - Persistent archive: Writes every brief to a dated, grep-able archive directory with an INDEX.md so past research compounds and stays searchable. - Use Case: Before a sales call, ask for a deep dive on a competitor and receive a cited brief covering their positioning, recent community sentiment, and open questions. ## Quick Start Ask the agent to do a deep dive on a topic, for example: research the current state of vector databases and tell me which one fits a small startup.
