deep-search

Conduct multi-round web research and generate a structured markdown synthesis report.

1.0k|75|Updated Mar 15, 2026
One-click install
npx skills add https://github.com/octos-org/octos --skill deep-search-octos-org
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: deep-search
Source: https://github.com/octos-org/octos/tree/main/crates/app-skills/deep-search
Command: npx skills add https://github.com/octos-org/octos --skill deep-search-octos-org

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Deep multi-round web research and structured synthesis across multiple sources save time and reduce manual guesswork when researching complex topics.

Core Features & Use Cases

  • Parallel web crawling: fetch multiple sources concurrently and surface diverse viewpoints.
  • Reference chasing: identify and retrieve highly cited external sources to anchor conclusions.
  • Structured synthesis: produce a report with sources, highlights, and a clear narrative across rounds.

Quick Start

Provide a topic to investigate and the tool will perform multi-round searches, parallel fetch pages, chase references, and generate a structured report in the research directory.

Frequently Asked Questions about deep-search

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate multi-round web research and generate a cited synthesis report?

Automate multi-round web research by providing a topic to trigger parallel fetching, reference chasing, and cross-round follow-ups, generating a cited markdown synthesis report and sources summary.

What is parallel web crawling for complex topic synthesis?

Parallel web crawling fetches multiple sources concurrently to surface diverse viewpoints, identifying and retrieving highly cited external references to anchor conclusions in a structured synthesis report.

Can I use deep search for topics requiring both breadth and depth?

Deep search handles topics requiring breadth and depth through multi-round searches, parallel page fetching, and reference chasing to build a comprehensive, cited narrative report.

Do I need special dependencies to run deep web research and crawling?

No special dependencies are required beyond access to configured search engines and a host-supplied synthesis configuration to execute deep web research and output a canonical markdown report.

How does reference chasing work in multi-round web research?

Reference chasing identifies and retrieves highly cited external sources during multi-round web research, anchoring conclusions and expanding the structured synthesis report with verified reference materials.

What is the best way to synthesize diverse viewpoints from crawled web sources?

Synthesize diverse viewpoints by parallel fetching multiple sources, chasing references across rounds, and generating a structured synthesis report with highlights, sources, and a clear narrative.