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

termyte

v0.2.7

Published

Termyte eats dangerous agent actions before they eat your codebase

Readme

Termyte MCP

Termyte is a minimum runtime safety layer for coding agents.

By default it gives agents one tool:

  • execute: run shell commands through Termyte governance and failure-memory checks.

Advanced context_build and guard_action tools are internal by default and can be exposed with TERMYTE_EXPOSE_ADVANCED_TOOLS=1 for custom integrations.

Why It Exists

Coding agents fail in two expensive ways:

  1. They run destructive actions too casually.
  2. They repeat failed approaches because they do not remember what went wrong.

Termyte blocks clearly destructive actions, warns on similar prior failures, redacts sensitive data, and stores sanitized runtime memory in the cloud so every authorized agent in the same organization can benefit from prior outcomes.

The cloud runtime also persists durable session state, so failure memory can surface the last relevant failure or warning without making the agent run a separate context step.

Install

npx termyte init

init installs MCP config for the detected agent and writes TERMYTE_PROTOCOL.md next to the agent config.

init also starts browser-mediated login. The terminal opens a Termyte activation URL, the user signs in on the website, and the CLI stores an organization-bound install token in ~/.termyte/config.json.

Free installs are limited to one active agent session. Team and enterprise organizations can run multiple agents against shared org-scoped memory.

Agent Protocol

Agents should follow this workflow:

  1. Use execute for shell commands.
  2. Treat WARN as proceed-with-caution and keep the warning in context.
  3. Treat BLOCK as do-not-proceed and explain the safer alternative.

Safe reads, tests, builds, and normal local edits are allowed. Sensitive surfaces warn. Irreversible actions such as secret access, destructive deletes, protected git mutation, package publishing, production database mutation, and production deploys block.

Verdicts

  • ALLOW: proceed.
  • WARN: proceed, but inject the warning or instruction patch into the agent context.
  • BLOCK: do not perform the action.

Cloud Data Posture

Termyte sends sanitized task, action, and outcome summaries to the cloud runtime by default. Redaction runs before transmission, judge input, persistence, logs, retrieval, and failure-memory storage.

Enterprise Controls

  • Org-scoped shared memory: no cross-organization retrieval.
  • Plan enforcement: free, team, and enterprise active-agent limits are enforced by the runtime.
  • Shadow mode: enterprise trials can observe would-block decisions before enforcement.
  • Audit timeline: GET /v1/governance/timeline reconstructs the causal sequence for demos, incidents, and compliance review.