npm package discovery and stats viewer.

Discover Tips

  • General search

    [free text search, go nuts!]

  • Package details

    pkg:[package-name]

  • User packages

    @[username]

Sponsor

Optimize Toolset

I’ve always been into building performant and accessible sites, but lately I’ve been taking it extremely seriously. So much so that I’ve been building a tool to help me optimize and monitor the sites that I build to make sure that I’m making an attempt to offer the best experience to those who visit them. If you’re into performant, accessible and SEO friendly sites, you might like it too! You can check it out at Optimize Toolset.

About

Hi, 👋, I’m Ryan Hefner  and I built this site for me, and you! The goal of this site was to provide an easy way for me to check the stats on my npm packages, both for prioritizing issues and updates, and to give me a little kick in the pants to keep up on stuff.

As I was building it, I realized that I was actually using the tool to build the tool, and figured I might as well put this out there and hopefully others will find it to be a fast and useful way to search and browse npm packages as I have.

If you’re interested in other things I’m working on, follow me on Twitter or check out the open source projects I’ve been publishing on GitHub.

I am also working on a Twitter bot for this site to tweet the most popular, newest, random packages from npm. Please follow that account now and it will start sending out packages soon–ish.

Open Software & Tools

This site wouldn’t be possible without the immense generosity and tireless efforts from the people who make contributions to the world and share their work via open source initiatives. Thank you 🙏

© 2026 – Pkg Stats / Ryan Hefner

research-agent-example

v0.1.0

Published

Autonomous research agent — ReAct strategy, DuckDuckGo HTML scrape (no API key)

Readme

Autonomous Research Agent

An autonomous research agent that uses the ReAct (Reasoning + Acting) strategy to iteratively search the web, scrape pages, extract facts, and synthesize a comprehensive report — with no search API key required.

Pipeline Graph

ReAct Loop (up to N iterations):
  THINK → what to search/scrape/extract next
  ACT   → one of: search-web | scrape-page | extract-facts | synthesize-report
  OBSERVE → feed result back to agent for next decision
  ↓
[when agent decides it has enough info]
synthesize-report → structured report with findings + sources + gaps

Features

  • No API key for search — uses DuckDuckGo HTML endpoint
  • Autonomous — agent decides what to search and when to stop
  • ReAct strategy with periodic reflection to avoid loops
  • Structured output — title, summary, findings, sources, confidence, gaps
  • Trace logging — see every thought and action the agent took

Setup

cp .env.example .env
# Fill in ANTHROPIC_API_KEY (or OPENAI_API_KEY)
pnpm install
pnpm build

Usage

# Research a topic
node dist/src/run.js --topic "The impact of LLMs on software development"

# More iterations for deeper research
node dist/src/run.js --topic "Quantum computing 2024 state" --max-iter 12

# Save report
node dist/src/run.js --topic "Climate tech startups 2024" --output ./report.md

Environment Variables

| Variable | Default | Description | |----------|---------|-------------| | ANTHROPIC_API_KEY | — | Anthropic API key | | OPENAI_API_KEY | — | OpenAI API key (alternative) | | OPENCODE_BASE_URL | — | Use local OpenCode proxy | | MAX_ITERATIONS | 8 | ReAct loop max iterations | | MAX_SOURCES | 5 | Max pages to scrape |

How ReAct Works

The agent follows a loop:

  1. THINK: The LLM reasons about what to do next given all previous observations
  2. ACT: Execute one skill (search, scrape, extract, or synthesize)
  3. OBSERVE: Feed the result back; the LLM sees what the skill returned
  4. Repeat until the agent calls synthesize-report or hits MAX_ITERATIONS

Every 4 steps, the agent reflects on its progress to avoid getting stuck in loops.