What is the diffusion research pattern?

The diffusion research pattern is an iterative process that starts with a noisy or speculative understanding and refines it through parallel, isolated workers to ensure diverse perspectives and minimize bias.

Does it provide sources for its findings?

Absolutely. Every final report includes a complete source list with URLs and descriptions for all factual claims, ensuring all information is verifiable.

Can I run this research in the background?

Yes, the skill supports a background execution mode via an orchestrator subagent, allowing you to continue other work while the research is performed autonomously.

How many iterations does it perform?

The skill performs a maximum of three iterations to prevent infinite loops while ensuring a balance between comprehensive depth and execution efficiency.

How does the skill ensure research quality?

It uses a separate gap-detector agent to objectively evaluate coverage against the research brief and triggers up to three iterations to fill any missing information.

Deep Research Diffusion

Name: Deep Research Diffusion
Author: RBozydar

byRBozydar

Web Scraping & Data Collection

Conducts exhaustive, multi-perspective research using parallel agent workers and iterative gap detection.

About

Deep Research Diffusion is a sophisticated Claude Code skill designed for high-stakes information gathering and structured analysis. It employs a multi-agent "diffusion loop" where independent workers investigate specific dimensions of a topic in isolation to prevent cognitive bias and herd behavior. The process includes an interactive refinement phase to align on goals, followed by autonomous worker execution and an objective gap-detection phase to ensure no stone is left unturned. By synthesizing findings from multiple sources with full citations, it transforms complex queries into professional-grade research reports.

Key Features

Interactive research brief refinement with multi-select user validation
Autonomous gap detection via objective third-party agent analysis
0 GitHub stars
Isolated worker execution to prevent information bias and ensure depth
Automatic synthesis of findings with comprehensive source citations
Parallel multi-agent research workers for diverse perspectives

Use Cases

Investigating emerging trends in scientific fields like quantum computing or drug discovery
Conducting thorough competitive analysis or AI governance framework reviews
Comparing complex architectural approaches or technical frameworks across industry providers

About

Key Features

Interactive research brief refinement with multi-select user validation
Autonomous gap detection via objective third-party agent analysis
0 GitHub stars
Isolated worker execution to prevent information bias and ensure depth
Automatic synthesis of findings with comprehensive source citations
Parallel multi-agent research workers for diverse perspectives

Use Cases

Investigating emerging trends in scientific fields like quantum computing or drug discovery
Conducting thorough competitive analysis or AI governance framework reviews
Comparing complex architectural approaches or technical frameworks across industry providers