Slapify

Slapify is a fully autonomous browser agent. Give it a goal in plain English and it does what a person at a keyboard would: reads pages, clicks, types, logs in, fills forms, uploads and downloads files, works across tabs, and keeps going until the goal is done. A run can last a few seconds or several days: the agent can schedule itself, sleep between cycles, remember what it has done, and report to you on Telegram.

The same package also runs plain-English end-to-end test flows (.flow files) and performance audits, from the CLI or from Node.js.

It is built to be left alone. Every action has a contract, irreversible actions (posting, sending, buying) can be reviewed against your own policy before they run, repeated failures get guidance instead of dead ends, and an LLM outage makes the run wait rather than exit.

Install

npm install -g slapify

# or run without installing
npx slapify init

Requires Node.js 18 or later. The current release on npm is 0.0.23.

Quick start

# 1. Create .slapify/config.yaml (pick a provider, a model and a browser)
npx slapify init

# 2. Give the agent a goal
npx slapify task "Find three well-reviewed ramen places near Union Square open after 10pm and put them in ramen.json"

From code:

import { createAgent } from "slapify";

const agent = createAgent({ stealth: true });
const run = await agent.run("Check the status of my last three orders on example-shop.com");
console.log(run.finalSummary);

Modes

Two more commands: slapify fast lets a decision model drive simple click-and-fill goals in seconds, and slapify server keeps an agent online that takes new goals from Telegram.

Run it on a subscription or any API

llm can be a list. When one model is rate-limited or out of quota, the agent moves to the next, and when every model is down it waits and retries. See Models and subscriptions.

Documentation