deep-research

Generates multi-agent research reports on any topic via the DeepTutor CLI.

Updated Jul 10, 2026
One-click install
npx skills add https://github.com/AvaTar-ArTs/.Agent-skills --skill deep-research-avatar-arts
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: deep-research
Source: https://github.com/AvaTar-ArTs/.Agent-skills/tree/main/deep-research
Command: npx skills add https://github.com/AvaTar-ArTs/.Agent-skills --skill deep-research-avatar-arts

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires deeptutor.

What problem does it solve? Producing a thorough research report, literature review, or technology comparison normally requires manually gathering and synthesizing information from many sources. This Skill automates that process by invoking DeepTutor's multi-agent research pipeline from the command line. ## Core Features & Use Cases - Multiple Output Modes: Generate notes, full reports, comparisons, or structured learning paths depending on the goal. - Configurable Depth and Sources: Choose quick, standard, or deep research thoroughness, and draw from a knowledge base, the web, or academic papers. - Structured JSON Output: Returns parseable NDJSON so report content can be extracted and assembled programmatically. - Use Case: Ask for a deep comparison of agentic RAG versus traditional RAG using both your knowledge base and web sources, and receive a complete synthesized report. ## Quick Start Ask the agent to run a deep research report on your chosen topic, specifying the mode, depth, and sources you want.

Frequently Asked Questions about deep-research

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

FAQPage Schema
How do I generate a research report with DeepTutor?

Run deeptutor run deep_research with your topic, the --format json flag, and a --config-json object specifying mode, depth, and sources. All three config fields are required for the command to work.

What output modes does deep research support?

DeepTutor supports four modes: notes, report, comparison, and learning_path. Choose the mode in the config JSON based on whether you want a summary, full report, side-by-side comparison, or structured learning sequence.

Can I use my own knowledge base for deep research?

Yes, include kb in the sources array and pass the --kb flag with your knowledge base name. You can combine it with web or papers sources for hybrid research.

Why does deep research take so long to finish?

The deep depth setting runs a thorough multi-agent pipeline that can take several minutes. Set the exec tool timeout to 300 seconds or higher to avoid premature termination.

How do I parse the JSON output from deep research?

The output is NDJSON with one JSON object per line. Filter lines where type equals content and concatenate them to assemble the full report text.