# dzhng/deep-research

An AI-powered research assistant that performs iterative, deep research on any topic by combining search engines, web scraping, and large language models.  The goal of this repo is to provide the simplest implementation of a deep research agent - e.g. an agent that can refine its research direction overtime and deep dive into a topic.

Repository: https://github.com/dzhng/deep-research
Canonical: https://ross.abutalabs.com/products/deep-research
Language: TypeScript
License: MIT
License Family: permissive
Topics: agent, ai, gpt, o3-mini, research
Last push: 2026-04-11T23:58:25+00:00

## Health v2 (maintenance only)
Score: 55/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 76, release rhythm 35, longevity 41
- inputs: {"age_days": 576, "days_push": 144, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 19603, forks 1993 (observed 2026-08-28T04:11:28.955032+00:00)

## What it is
An AI-powered deep research agent that iteratively generates search queries, scrapes web content, and refines its research direction using LLMs to produce comprehensive markdown reports. It is intentionally kept under 500 lines of code to serve as a simple, hackable reference implementation.

## Use cases
- research any topic in depth with an AI agent
- generate a cited markdown research report from web sources
- automate iterative search query generation and refinement
- learn how to build a deep research agent
- deep dive into a topic with configurable breadth and depth

## When to choose
- you want a minimal, readable implementation of a deep research agent to build on
- you have Firecrawl and OpenAI API keys and want automated iterative web research
- you need a starting point for custom research automation in TypeScript

## When to avoid
- you need a production-grade research platform with UI, persistence, and team features
- you cannot use paid APIs like Firecrawl or OpenAI
- you need guaranteed factual accuracy without source verification

## Facets
- artifact type: application
- maturity: active
- function: agent-framework, web-scraping, search-engine, llm-inference, rag, prompt-engineering
- domain: artificial-intelligence, large-language-models, developer-tools
- platform: cli
- tags: deep-research, research-agent, firecrawl, openai, iterative-research, minimal-codebase, ai-agents, search, nodejs, typescript

## Member repositories
- dzhng/deep-research (main) score 55

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:11:28.955032+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T16:59:54.365606+00:00, confidence not recorded.
  - readme: https://github.com/dzhng/deep-research (fetched 2026-08-28T04:11:28.955032+00:00, sha 4daa965cfabb)
- Data as of 2026-08-30T08:39:29.467469+00:00.