How long does a research task take?

The skill uses hard timeouts to ensure speed: Quick mode takes 2 minutes, Standard takes 3 minutes, and Extensive mode takes up to 10 minutes.

Where are the research findings stored?

Findings are first stored in a timestamped scratchpad directory for drafting and then moved to a permanent history archive with a README and executive summary.

What happens if one research agent is slow?

The skill follows strict timeout rules. Once the time limit is reached, the main agent stops waiting and synthesizes the final report using all available data gathered up to that point.

How many agents can run simultaneously?

Depending on the mode, you can launch 3 agents (Quick), 9 agents (Standard), or up to 24 agents (Extensive) to cover different angles of a research topic.

What is the Multi-Agent Research Assistant skill?

It is a specialized capability for Claude Code that allows it to launch multiple parallel agents (Perplexity, Claude, and Gemini) to perform deep-dive research and synthesize findings into a single report.

Multi-Agent Research Assistant

Name: Multi-Agent Research Assistant
Author: rafaelcalleja

byrafaelcalleja

•

Web Scraping & Data Collection

Orchestrates parallel research agents across Perplexity, Claude, and Gemini to deliver synthesized, multi-perspective reports with source attribution.

The Research skill transforms Claude into a powerful information engine by deploying swarms of specialized agents to investigate topics simultaneously. By decomposing complex queries into sub-questions and running up to 24 parallel researchers, it provides comprehensive coverage that far exceeds standard search capabilities. The skill features three distinct modes—Quick, Standard, and Extensive—each with strict timeout rules to ensure rapid delivery of findings. Results are automatically structured, synthesized with confidence levels, and archived using a standardized history pattern for permanent reference.

Key Features

01Strict timeout management to ensure prompt synthesis of partial results

02Three research tiers: Quick (3 agents), Standard (9 agents), and Extensive (24 agents)

031 GitHub stars

04Automated query decomposition for multi-domain coverage

05Standardized workflow for scratchpad drafting and permanent history archiving

06Multi-engine orchestration using Perplexity, Claude, and Gemini agents

Use Cases

01Comprehensive technical investigations for documentation and project planning

02Rapid synthesis of current events from multiple academic and web sources

03Deep-dive analysis of emerging technologies and market trends

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add rafaelcalleja/claude-market-place research

For use in Claude.ai and ChatGPT

Download Skill