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

@eidentic/browser

v0.1.8

Published

Sealed browser-automation tools for Eidentic — first-class sandboxed browser tools over an injected Playwright-like page (PageLike interface).

Readme

@eidentic/browser

Sealed browser-automation tools for Eidentic — first-class sandboxed browser tools over an injected PageLike page. Pass any object that satisfies the PageLike interface: a real Playwright Page, or a faithful in-memory fake for tests.

Install

pnpm add @eidentic/browser playwright-core
npx playwright install chromium

Usage

import { chromium } from "playwright-core";
import { browserTools } from "@eidentic/browser";
import { Agent } from "eidentic";

const browser = await chromium.launch({ headless: true });
const context = await browser.newContext();
const page = await context.newPage();

const agent = new Agent({
  id: "web-agent",
  model,
  store,
  tools: browserTools(page, {
    allowlist: ["example.com", "docs.example.com"],
  }),
});

for await (const ev of agent.query("What is on the homepage of example.com?", { sessionId: "s-1" })) {
  if (ev.type === "stream.delta") process.stdout.write(ev.delta.text);
}

await browser.close();

Tools

| Tool ID | Side effect | Description | |---|---|---| | browser_navigate | destructive | Navigate to an http(s) URL. Validates before AND after navigation (redirect escape detection). | | browser_read | read-only | Read current page title, URL, and text (body or a CSS selector). Truncated to maxContentBytes. | | browser_click | destructive | Click an element by CSS selector. Errors are tool errors, not throws. | | browser_fill | destructive | Fill an input by CSS selector. Errors are tool errors, not throws. |

Security

Every browser_navigate call:

  1. Validates the target URL scheme (http/https only).
  2. Checks the host against the private-IP/loopback/metadata blocklist (SSRF defense, mirrors @eidentic/tools).
  3. Checks the host against your allowlist (when configured).
  4. After goto(), re-validates page.url() — so server-side redirects that would escape the allowlist or land on a private host are caught and returned as tool errors before the agent can act on them.

Options

browserTools(page, {
  // Optional: restrict navigation to these hostnames (and their subdomains).
  // Omit for no restriction; pass [] to deny all navigation.
  allowlist?: string[];

  // Default: true. Set to false only in controlled test environments.
  blockPrivateHosts?: boolean;

  // Max UTF-8 bytes in browser_read results. Default: 512 KB.
  maxContentBytes?: number;
})

PageLike interface

interface PageLike {
  goto(url: string): Promise<unknown>;
  content(): Promise<string>;
  innerText(selector: string): Promise<string>;
  click(selector: string): Promise<void>;
  fill(selector: string, value: string): Promise<void>;
  url(): string;
  title(): Promise<string>;
  screenshot?(): Promise<Uint8Array>; // optional, not exposed as a tool in v1
}

A real playwright-core Page satisfies this interface structurally. No adapter needed.

Roadmap

  • browser_screenshot: returns a base64-encoded screenshot string. Not in v1 because binary results don't compose cleanly with text tool results; the encoding overhead and context-window cost warrant a dedicated design.

Links

Apache-2.0