LoopEngineBETA

Docs

Documentation

How LoopEngine works, from scaffold to durable approvals — agents, adapters, permission rules, and abilities.

Installation

LoopEngine is a runtime for building AI agents: a persona, a set of tools, and permission rules, run through a transparent ReAct loop. No chain DSL, no hidden control flow — core/run-agent.ts is a single function you can read top to bottom.

There are two ways to get it running, depending on what you're doing.

Scaffold a new project

This is the intended way to use LoopEngine — an independent project you own and build agents in, not a library you import into an existing app:

npx create-loopengine@latest my-agents
cd my-agents
npm install
cp .env.example .env   # fill in ANTHROPIC_API_KEY
npm run dev             # HTTP server on :8787 — same as `npx loopengine dev`

This scaffolds a standalone project — its own repo, with loopengine as a dependency, a starter agent, and CLI + HTTP adapters already wired up. See Directory Structure for what that layout looks like, and Add an Agent for defining your first real agent.

Clone the repo directly

For exploring the runtime itself or its own reference agents (customer-service, file-agent, rag-agent, support-orchestrator, web-agent) rather than scaffolding a new project:

git clone https://github.com/loopengine-co/loopengine.git
cd loopengine
npm install
echo "ANTHROPIC_API_KEY=sk-..." > .env
npx tsx adapters/http.ts   # HTTP server on :8787, no build step

npx tsx runs the TypeScript source directly, same as the scaffolded project's own npm run dev. For a built, production-style run instead:

npm run build
npm start   # node --env-file-if-exists=.env dist/adapters/http.js

What's next

Open http://localhost:8787/agents to see it running — the Web UI page covers the Admin UI and playground you'll land on from there. From here: Directory Structure, then Basic Configuration, then Add an Agent.