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

@researchlog/app

v0.3.1

Published

App-driven transparent research log: an agent does the research, you judge it.

Readme

Research Log

A transparent research log: you create a topic, an agent does the web research and records every source it opens — URL, what it looked for, a neutral summary, separate relevance/credibility ratings (each with a rationale), and an honest judgement. You review, comment, and re-run; the agent reads your comments and continues.

Quick start

npx @researchlog/app

Then open http://127.0.0.1:3100.

That's the whole app — REST API, the agent's MCP server, the live WebSocket feed, an embedded database, and the web UI — in one local process bound to loopback.

Prerequisite: the claude CLI

Real research runs spawn your local claude CLI, so it must be installed and authenticated (claude on your PATH). Runs use your own subscription quota. Use the "Check Claude Code" button in the top bar to confirm the CLI is installed and signed in before you start a run. To try the app without an LLM, use the scripted adapter:

CONRES_ADAPTER=fake npx @researchlog/app

Where your data lives

A ./.research-log/ directory is created in whatever folder you run the command from — it holds the entire database. Delete it to reset; back it up to keep your research. Point it elsewhere with CONRES_DATA_DIR.

Configuration

| Variable | Default | Purpose | |---|---|---| | CONRES_PORT | 3100 | Server port | | CONRES_HOST | 127.0.0.1 | Bind host (loopback). Non-loopback is refused unless auth is on. | | CONRES_DATA_DIR | ./.research-log/pglite | Database directory (memory:// for in-memory) | | CONRES_ADAPTER | claude_local | claude_local (real CLI) or fake (no LLM) | | CONRES_CLAUDE_BIN | claude | Path to the claude binary | | CONRES_AUTH_MODE | local_trusted | local_trusted (no login, loopback-only) or authenticated |

Exposing it past your own machine

By default the server binds to loopback and refuses to bind a public host — so an open board is never exposed by accident. To expose it (LAN, tunnel, etc.), turn on auth:

CONRES_AUTH_MODE=authenticated \
CONRES_AUTH_PASSWORD_HASH=... \
CONRES_SESSION_SECRET=... \
CONRES_HOST=0.0.0.0 \
npx @researchlog/app

A single board password (gate every request via a signed, httpOnly session cookie) is the one human-facing credential. The agent's MCP write endpoint is always loopback-only, regardless of mode. Set a stable CONRES_SESSION_SECRET so sessions survive restarts.