Rule-gated automatic approval for DeepSeek Harness sandbox escalations with an optional LLM reviewer and native human fallback.

Model-based permission approval (approve-for-me) for DeepSeek Harness: an approval/request answerer backed by a separate reviewer model

Secondary-model review of model-issued Pi tool calls

node

Let agents run disposable VMs

Web ask_user_question feature: host tool mount plus composer-takeover question UI

Help command for node, partner of minimist and commist

Let your JS API users either give you a callback or receive a promise

A low level parser for ANSI sequences.

A replacement for process.exit that ensures stdio are fully drained before exiting.

A package to run an embedded Postgresql database right from NodeJS

The TypeScript types for the hfs project.

Match human-quality input to potential matches by edit distance.

A high level git url parser for common git providers.

Abstract user-questions seam (ctx.userQuestions) for asking the human during agent runs

An advanced url parser supporting git urls too.

Intuitive, type safe and flexible Store for Vue

.properties file parser, editor, formatter and bundler integrations.

React components for efficiently rendering large, scrollable lists and tabular data

react native in app review, to rate on Play store, App Store, Generally, the in-app review flow (see figure 1 for play store, figure 2 for ios) can be triggered at any time throughout the user journey of your app. During the flow, the user has the ability