Open source browser automation tools, installed and run

Browser automation tools you can run yourself. The list is ordered by whether each project came up when Argusic installed it fresh, with the recording of that attempt one click away.

Tested between and . Each row shows its own test date; a project can change after that day.

104 of 107 tested projects run. 4 more waiting for a test.

In short: 70 of the 107 tested projects started as-is on a fresh machine: browser-harness, steel-browser, ferret, playwright-go, browser-control, webcmd, BrowserKitten, and Mink, and 62 more. 34 more started once a stand-in replaced a service they expect, such as a database: Jobs_Applier_AI_Agent_AIHawk, ego-lite, GenericAgent, llm-scraper, google-qr-sms-helper, brightdata-mcp, lcu, and actionbook, and 26 more. 3 could not be verified: OpenBot, browsentic, and pagesift; the log shows where each one stopped.

Measured by Argusic on a fresh machine every time. Every number links to its evidence.

#projectverdictArgusic Scorelanguagestarstested on
1browser-harnessRuns100 / 100Python19,193

Let Claude Code, Codex or Cursor use your real Chrome, logged in as you. One prompt to set up. Open source, MIT.

What the test found: browser-harness 0.1.13 installs from source, all 266 unit tests pass, and the CLI controls a headless Chromium via CDP, page_info, navigation, tab management, JS execution, and screenshots all work end-to-end. 12 minutes.

2steel-browserRuns100 / 100TypeScript7,999

🔥 Open Source Browser API for AI Agents & Apps. Steel Browser is a batteries-included browser sandbox that lets you automate the web without worrying about...

What the test found: With a user-local Node 22 and Chrome for Testing plus DISABLE_CHROME_SANDBOX=true, the dev-mode API runs on port 3000, /v1/health returns 200 with live Chrome, and sessions/scrape/screenshot/pdf plus external CDP client connections all worked with real results. 11 minutes.

3ferretRuns100 / 100Go6,025

Declarative data automation language and Go runtime for structured extraction workflows.

What the test found: Ferret v2 repo builds with Go 1.25.0, all 55 test packages pass (unit, integration, security, spec, compat, benchmarks), and the CLI binary correctly evaluates FQL queries (return 1 + 1 → 2). 5 minutes.

4playwright-goRuns100 / 100Go3,547

Playwright for Go a browser automation library to control Chromium, Firefox and WebKit with a single API.

What the test found: Go! v1.22.2 toolchain works, Playwright driver v1.62.1 installed, projects builds with 'go build .', 85 unit tests pass, full integration test suite passes with real Chromium/Firefox/WebKit browser automation. 30 minutes.

5browser-controlRuns100 / 100Rust3,160

A tiny, fast Rust CLI that drives a real browser over the Chrome DevTools Protocol, built for coding agents.

What the test found: browser-control-cli v0.1.1 builds, all 10 unit tests pass, formatting check passes (verify.sh), and the binary launches Chrome headlessly, navigates to URLs, evaluates JS, snapshots page elements, clicks element refs, and issues raw CDP commands, all verified end-to-end with Chromium 119 against https://example.com. 15 minutes.

6webcmdRuns100 / 100TypeScript2,795

Self-learning agent browser

What the test found: npm install and TypeScript build succeed, 3571 unit tests pass with 0 failures, and webcmd doctor verifies the daemon, CloakBrowser binary, and network connectivity are all operational. 2 minutes.

7BrowserKittenRuns100 / 100JavaScript2,402

A Chrome side-panel agent for your signed-in browser. Read pages, take action, create files, and run local follow-ups. Bring your own models.

What the test found: Browser Kitten v1.2.0 source loads, all 813 unit tests pass with Node.js 22, and the release packaging pipeline produces a verified extension/ directory with 273 files and matching source parity. Requires Node.js 22+ to run tests and Chrome 135+ at runtime. 5 minutes.

8MinkRuns100 / 100PHP1,605

PHP web browser emulator abstraction

What the test found: PHP 8.3.6 extracted from Ubuntu 24.04 debs, Composer 2.11.2 installed, 39 PHP dependencies installed, phpunit 9.6.39 runs and passes all 520 tests with 1165 assertions. 5 minutes.

9AIPexRuns100 / 100TypeScript1,285

AIPex: AI browser automation assistant, no migration and privacy first. Alternative to Manus Browser Operator、 Claude Chrome and Agent Browser

What the test found: The project builds, typechecks, and all 639 tests pass across 5 packages using Node 22.15.0. 4 minutes.

10dsh-browserRuns100 / 100TypeScript1,119

Chrome sidebar extension that lets DeepSeek Harness operate your browser directly, no vision capabilities required. 一款 Chrome 侧边栏扩展程序,可让 DeepSeek Harness...

What the test found: Bridge plugin and Chrome MV3 extension build, typecheck, and both test suites pass (14+50 suites, 140+388 tests). The dsh web CLI cannot boot due to an upstream exitOverride Commander bug in @deepseek-ai/dsh 0.1.5-rc.2. 8 minutes.

11beachpatrolRuns100 / 100JavaScript925

🏝️ A CLI Tool to Automate Your Everyday Web Browser.

What the test found: Beachpatrol installs and runs: npm install + playwright install chromium succeed, beachpatrol --help prints usage, beachpatrol --version prints v0.0.6, and the full E2E test suite (14 tests) passes 13/14 with zero failures, exercising smoke test, error propagation, plain/dep/download commands, concurrent routing... 4 minutes.

12puppetryRuns100 / 100JavaScript888

Web testing solution for non-developers on top of Puppeteer and Jest

What the test found: npm install succeeds, webpack build succeeds, all 21 unit test suites pass (261 tests), and the Electron app launches on the virtual display logging renderer initialization. 21 minutes.

13Prism-Browser-CommunityRuns100 / 100TypeScript769

Local-first open-source multi-profile fingerprint browser based on Chromium and Electron

What the test found: Project builds (typecheck + electron-vite build), all 126 unit tests pass, the Electron app launches on a virtual display and stays up, and the kernel patch lock file is fully validated. 2 minutes.

14agenticosRuns100 / 100Python716

Open-source, self-hosted AI agent platform for your company's AI teammates. Build, run and govern agents on your own servers and models: shared chat, RAG...

What the test found: Backend Python environment (Python 3.15.2 via uv) installed, 254 packages resolved; frontend Node dependencies (771 packages) installed via bun; backend lint passes (ruff, formatting); 609 backend unit tests pass covering coverage gate, agent registry, labels, A2A client/server, mutation locks, observability, and... 15 minutes.

15ocRuns100 / 100JavaScript694

Turn any website into a compact CLI tailored for AI agents. Browse the web in hundreds of tokens, not tens of thousands.

What the test found: only-cli v0.5.8 installs and runs on Node 20.11.1; 263 of 264 tests pass; the CLI fetches, distills, and renders public web pages with token-budgeted output and actionable numbered links. 3 minutes.

16scoutRuns100 / 100TypeScript673

Scout: the open-source browser agent, by Outcrawl. Mirrored automatically from Outcrawl's own repository.

What the test found: Bun 1.4.3 and Node 22.14.0 installed locally, bun install succeeded, and every workspace test suite passes (1453 pass, 0 fail, 0 skip) including the previously-skipped real headless-Chromium tests; a standalone Scout run completed with status completed and data 'All stories loaded.'. 10 minutes.

17WebsiteBenchRuns100 / 100Python640

Can AI agents rebuild websites through browser exploration?

What the test found: Full Python environment installed with uv; all 54 unit tests pass; ruff linter clean; all 70 task definitions pass static validation; websitebench CLI responds with version and help, lists all 70 tasks, and downloads real tasks from Harbor Hub without any credentials. 3 minutes.

18huashu-chromeRuns100 / 100JavaScript610

让任何 AI agent 操控你自己的 Chrome, 带着全部登录态。MCP + Chrome 扩展,Claude Code / Codex / Cursor / Gemini CLI 通用。

What the test found: All 123 unit tests pass, npm install completes, MCP server responds with 22 tools, bridge daemon starts and answers probes, doctor command runs successfully. 2 minutes.

19readerRuns100 / 100TypeScript579

Open source web infrastructure for AI. Scrape, crawl, and automate the web, clean markdown, browser sessions, ready for your agents.

What the test found: Reader CLI builds, all 346 unit tests pass, the daemon starts with Playwright's bundled Chromium, and an end-to-end scrape of https://example.com via HTTP returns clean HTML and markdown content with a 200 status. 13 minutes.

20bladebroRuns100 / 100Rust474

A Fully free agentic browser driver for AI , few tools, full control, real stealth, top-tier token efficiency.

What the test found: Bladebro builds from source with zero clippy warnings, all 217 tests pass, Chrome for Testing launches under Xvfb, and the binary navigates real websites, reads page content, and queries cookies. 8 minutes.

21wmuxRuns100 / 100TypeScript459

The ADE for AI agents. Automate everything else. Do your true work.

22html2imageRuns100 / 100Python455

A package acting as a wrapper around the headless mode of existing web browsers to generate images from URLs and from HTML+CSS strings or files.

What the test found: Package installed in venv, ChromiumHeadless defaults include --no-sandbox, 17 Chrome tests pass, CLI produces correct images, Edge tests gracefully skipped. 5 minutes.

23camofox-browserRuns100 / 100JavaScript453

Anti-detection browser server for AI agents, REST API wrapping Camoufox engine with OpenClaw plugin support

What the test found: Camofox Browser Server builds, starts on port 9998, creates browser contexts via Camoufox Firefox headless, navigates to https://example.com, returns accessibility snapshots, and passes the 677-test full suite with Node.js 20 and rebuilt native addons. 34 minutes.

24reinsRuns100 / 100Python394

Stop AI agents from doing things you didn't ask for.

What the test found: npm install and tsc build succeed, all 50 test cases pass, CLI help responds with all commands listed. 17 minutes.

25ash-agentsRuns100 / 100Python390

A swarm of disposable agents for unattributable web interaction

What the test found: Ash Agents orchestrator server runs on port 8092 with ASH_DEMO_MOCK=1, all endpoints respond correctly, and all 12 unit tests pass. 6 minutes.

26dsh-deepseek-web-loginRuns100 / 100JavaScript382

Unofficial DSH (DeepSeek Harness) plugin: use chat.deepseek.com web models as an LLM provider - browser-login capture, PoW solving, SSE streaming...

What the test found: The project builds cleanly (tsdown produces lib/index.js and lib/client.js), all 62 offline test files pass, and the bundle product verification confirms all runtime features are present. 5 minutes.

27jev-job-searchRuns100 / 100TypeScript378

Find software internships and new-grad jobs, fill the application forms, check every answer on the page, and send the ones you approve. JEV decides, code acts...

What the test found: The project installs and builds on Node.js v22.23.4; 420 unit/integration tests pass covering CLI commands, form mapping, job filtering, CSV, dates, caching, memory, mail, accounts, knowledge, documents, JEV client, pacing, templates, and more, only the 6 browser-level tests are skipped due to Chrome being unavailable... 2 minutes.

28pi-desktopRuns100 / 100TypeScript355

A local-first, cross-platform Electron desktop workspace for Pi Coding Agent, with sessions, project files, browser tools, skills, plugins, and messaging...

What the test found: Pi Agent Desktop v0.3.0 builds, all 1106 tests pass, and Electron smoke test verifies agent-host (pi=0.87.1) with all 12 function domains (renderer, RPC, session, worktree, git, watch, download, skills, channels, chat-appearance, toolchain) passing. 11 minutes.

29chrome-agentRuns100 / 100Python310

CLI that gives AI coding agents direct access to Chrome via the DevTools Protocol. Many agents and humans can share one browser, each with isolated event...

What the test found: chrome-agent 0.8.2 installed via pip, launches Chrome 157 headless with --no-sandbox, all 237 tests pass against a real browser over CDP. 12 minutes.

30webframeRuns100 / 100Go292

Run a web page like a program

What the test found: webframe builds from source with patched glaze (preferGTK3=true), runs with WebKitGTK 4.1 from deb packages, executes JS pages, reports console output, supports exit codes and scripts. 19 minutes.

31teracrawlRuns100 / 100TypeScript291

High-performance web crawler API optimized for LLMs. Turn any search or website into clean Markdown using remote browsers. Firecrawl alternative

What the test found: Teracrawl server starts on port 8085 with a local Chromium-backed mock Browser.cash pool and a mock SERP service. GET /health returns 200, POST /scrape converts a live URL to markdown via real headless Chromium, POST /serp/search returns mock search results, and POST /crawl combines both to scrape multiple URLs. 15 minutes.

32advai-cliRuns100 / 100Python283

Local-first CLI for agent skills, browser automation, knowledge bases, external CLIs, and terminal AI chat.

What the test found: advai-cli 1.0.12 installed from source, all 78 unit tests pass, the CLI responds to all top-level commands, and the knowledge base subsystem creates, adds, searches, and syncs real local files. 1 minute.

33advai-cliRuns100 / 100Python283

Local-first CLI for agent skills, browser automation, knowledge bases, external CLIs, and terminal AI chat.

What the test found: advai-cli 1.0.12 is installed in a venv, all 78 tests pass, and core commands (info, update, skill, kb, browser, cli) respond with valid output. 1 minute.

34openbrowser-aiRuns100 / 100Python279

OpenBrowser is a framework for intelligent browser automation. It combines direct CDP communication with a CodeAgent architecture, where the LLM writes Python...

What the test found: All 8030 unit and integration tests pass, CLI prints version and help, and headless Chromium via CDP navigates to about:blank and prints output correctly. 27 minutes.

35codexqaRuns100 / 100Python278

codexqa: 11 local-first Agent Skills for Cursor, Claude Code, Codex & OpenClaw, change impact analysis, AI code review, defect scan, testcase generation...

What the test found: The repo installs from a local checkout via npx skills with Playwright Chromium and the @openqa-cn/codexqa CLI in a user prefix; all repository checks pass: check-docs, compose-release-notes, checkout-boundary and inventory-service verifiers, polyglot fixture build, code-wiki 47/47, defect-analyzer full suite... 6 minutes.

36WeWorkerRuns100 / 100JavaScript266

Open-source desktop AI agent that takes real actions on websites, 33 adapters for mail, shopping, social, and more.

What the test found: Project installs cleanly, builds all artifacts, runs 1779/1779 webcli + 608/611 weworker tests (3 cancelled from pre-existing async flakiness); 41 backend-ts tests fail solely due to missing SQLite FTS5 in the pre-built Node.js 22 Linux binary. 16 minutes.

37agent-workspace-linuxRuns100 / 100Rust230

Isolated Linux desktop workspaces for AI agents, a hidden, agent-owned desktop and browser over MCP, so an agent can do GUI and web work without touching your...

What the test found: agent-workspace-linux 0.6.2 builds from source, all 199 tests pass, binary runs --help and --version, starts an X11 workspace with openbox and xdotool (extracted from apt debs), launches GUI apps (xlogo), takes screenshots via ffmpeg, and responds to MCP protocol initialization. 14 minutes.

38Uni-CLIRuns100 / 100TypeScript229

One command for every interface, search, run, and inspect real software across APIs, browsers, desktops, local tools, and MCP.

What the test found: Uni-CLI v1.2.1 builds from source, runs 3735 passing unit tests, and executes live commands (hackernews top, search, doctor) against real remote APIs on Node 22.23.4 (npm v10.9.9). The npm global package v1.3.0 is also installed and functional. 20 minutes.

39blueboxRuns100 / 100Python203

Index the world's undocumented APIs

What the test found: All 2434 unit tests pass after fixing 2 flaky async tests; pip editable install succeeds; CLI help commands work for bluebox-agent and bluebox-discover; core Python imports resolve without errors. 3 minutes.

40sample-strands-agent-with-agentcoreRuns100 / 100Python203

Reference architecture for agentic AI chatbots with Strands Agents and Amazon Bedrock AgentCore

What the test found: Python backend (FastAPI on port 8080) installs, starts, passes all 914 tests, and responds 200 on /health and /ping; browser tool (nova_act) and frontend (Next.js 16 requiring Node 20+) are unavailable due to container environment limitations. 6 minutes.

41agent-chrome-relayRuns100 / 100Go177

Let your AI agents use your Chrome, signed in as you, in background tabs. By Alexey Indeev · blog: read.aindeev.com · X: @AlexeyIndeev

What the test found: The Chrome Relay relay server starts, answers HTTP endpoints with correct status codes, and passes all 12 integration tests including agent WebSocket connections authenticated by secret and process markers, extension origin validation, web page rejection, and audit trail persistence. 20 minutes.

42invisible_puppeteerRuns100 / 100Python156

invisible puppeteer: Puppeteer for Python that anti-bots cannot see, so no captchas: same API, anti-detect stealth headless Firefox, undetected fingerprint...

What the test found: Editable install resolved all dependencies, default pytest suite passes (219/259), CLI version and fetch commands work, and the browser engine was downloaded and cached. 3 minutes.

43invisible_dotsRuns98 / 100TypeScript58

invisible dots: open-source, self-hosted alternative to OpenAI Dots, Grok Bot. Built to be undetectable by anti-bot systems.

What the test found: npm ci installed 428 packages, the CLI bundle and web client built, npm run typecheck passed, npx vitest run gave 2103 passed/3 skipped and go test gave 120 passed after I fixed the secret-input backspace bug and the execEnv DISPLAY bug, and the control plane served HTTP 200 on /api/health, /api/dots and the web UI... 22 minutes.

44camofox-browserRuns96.7 / 100JavaScript12,456

Stealth headless browser for AI agents, bypass Cloudflare, bot detection, and anti-scraping. Drop-in Puppeteer/Playwright replacement.

What the test found: Camofox browser server installed and verified: Node 22.23.4 in user space, Camoufox 1.3GB binary downloaded, GeoIP 50MB database downloaded, server starts at localhost:9377 with Xvfb virtual display, creates browser tabs navigating to real URLs, returns accessibility snapshots with element refs, and closes cleanly on... 13 minutes.

45mcpRuns96.7 / 100TypeScript7,402

Browser MCP is a Model Context Provider (MCP) server that allows AI applications to control your browser

What the test found: The repo now installs and builds under Node 24 + pnpm in this container, its full unit suite passes (252 tests green), the built daemon answers /browsermcp/status with 200, and the pnpm smoke suite passes all 18 sections on its good runs while intermittently flaking in headless-Chromium container conditions. 37 minutes.

46dotsRuns96 / 100Python32,611

AIHawk: AI job application bot, automated job search, job hunting, auto-apply to resume-matching jobs while you sleep.

What the test found: dots installs, its 4 tests pass, and the UI server runs and drives its bundled browser with a real model call, navigating https://example.com and reporting HTTP 200. 7 minutes.

47Jobs_Applier_AI_AgentRuns96 / 100Python32,611

AIHawk: AI job application bot, automated job search, job hunting, auto-apply to resume-matching jobs while you sleep.

What the test found: All 211 tests pass and the CLI responds correctly to --help and --version. The project is ready for use with an OrcaRouter API key. 3 minutes.

48page-agentRuns96 / 100TypeScript29,734

JavaScript in-page GUI agent. Control web interfaces with natural language.

What the test found: Install, sequential build, typecheck, lint, and all 69 unit tests pass. The project is fully functional with Node v22. via nvm. 27 minutes.

49atomic-agentRuns96 / 100TypeScript3,588

Atomic Agent is a local-first AI agent. Runs open-weight models on your own machine via llama.cpp.

What the test found: Build compiles, CLI binary runs and responds to --version/--help, 594 of ~600+ test files pass with 1 genuine test-logic failure in sidecar send-message concurrency, and all 22 eval-memory test files pass. 58 minutes.

50js-reverse-mcpRuns96 / 100TypeScript3,085

AI Agent-first JS 逆向 MCP Server:有头 Chrome 调试、断点、网络/WebSocket 分析、Patchright 反检测,可选 CloakBrowser。

What the test found: Project builds, all 19 unit tests pass, routing validation passes, and the MCP server responds to JSON-RPC tools/list requests. 3 minutes.

51openserpRuns96 / 100Go1,608

Self-hosted web search for AI agents and SEO. Google, Bing, Yandex, Baidu, DuckDuckGo & Ecosia, rendered in a real browser: ranked results and clean page...

What the test found: OpenSERP builds from source, all 12 unit test suites pass, the HTTP server serves /health (200) and handles live DuckDuckGo searches (200) returning structured JSON with organic results and SERP features. 6 minutes.

52WebAI2APIRuns96 / 100JavaScript1,428

WebAI2API: 基于 Camoufox 的网页 AI 转 API 工具,支持 LMArena/Gemini等,多窗口并发与账号隔离。 | Web AI to OpenAI API via Camoufox. Supports LMArena/Gemini and more, multi-window...

What the test found: The server starts, launches a Camoufox browser instance in headful mode on Xvfb, serves the OpenAI-compatible API on port 3000 returning 200 for /v1/models and WebUI, the SQLite history database initializes, all 19 adapters load, and the browser fingerprint is generated and saved. 6 minutes.

53Gemini-NexusRuns96 / 100JavaScript1,131

Gemini Nexus 是一款面向浏览器场景的 AI 助手扩展,集成 Gemini Web、Gemini API 与 OpenAI 兼容接口,支持网页上下文、图像处理、工具调用和 MCP 浏览器控制。

What the test found: Dependencies installed, Vite build compiles 1790 modules without error, all 1437 tests in 201 test files pass, and the Chrome extension packages successfully to artifacts/chrome-extension/. 7 minutes.

54auto-browserRuns96 / 100Python929

Give your AI agent a real browser, with a human in the loop. Open-source MCP-native browser agent.

What the test found: The controller package installed, all 1189 host-side tests pass, lint is clean, and the app starts and serves a real browser session over MCP+CDP, successfully navigating to example.com. 10 minutes.

55rustwrightRuns95 / 100Python962

Playwright's API on a Rust CDP engine, Chromium browser automation for Python & Node, no driver subprocess. (Alpha)

What the test found: Rustwright 0.3.0 Python package built from source, Chromium installed at /home/runner/.cache/ms-playwright/chromium-158.0.8127.36/chrome-linux64/chrome, smoke example prints 'Rustwright works', cargo check --locked passes. 17 minutes.

56pinchtabRuns94.3 / 100Go10,603

High-performance browser automation bridge and multi-instance orchestrator with advanced stealth injection and real-time dashboard.

What the test found: A source-built PinchTab server (binary reports pinchtab dev) is running on 127.0.0.1:9867 with a responsive always-on headless instance and answered authenticated health, navigate, snapshot, text, title, and screenshot API calls with HTTP 200, while the full Go unit suite (9166 passed, 18 skipped) and scripts/check.sh... 19 minutes.

57scrapy-playwrightRuns93.5 / 100Python1,470

🎭 Playwright integration for Scrapy

What the test found: scrapy-playwright 0.0.48 is installed with Scrapy 2.20 and Playwright 1.65; Chromium browser launched and renders pages through the Scrapy download handler; 47 Chromium tests and 29 unit tests pass. Firefox and Webkit browsers not installed, remote-browser tests need unavailable external processes. 12 minutes.

58graspRuns92 / 100TypeScript368

Grasp helps you manage multiple projects and parallel coding agents in one visual workflow, so humans can understand faster and ship more.

What the test found: With user-local Go 1.27.3 and Node 24.12.0, the built Grasp server and gateway binaries start and answer health 200, admin login and project APIs return 200 with seeded data, and the server, web, docs, gateway, and sandbox build, test, and coverage suites all pass. 18 minutes.

59ArchiveBoxRuns91.3 / 100Python29,077

🗃 Open source self-hosted web archiving. Takes URLs/browser history/bookmarks/Pocket/Pinboard/etc. saves HTML, JS, PDFs, media, and more...

What the test found: ArchiveBox 0.9.65rc24 installed in a Python 3.13 venv; data directory initialized with SQLite; Django admin server starts and serves HTTP on port 5797; 125 tests pass; full archiving (add, crawl) requires wget/chrome/yt-dlp which are not installed in this container without root. 12 minutes.

60nanobrowserRuns90.7 / 100TypeScript14,368

Open-source AI agent for web automation. Use your real, logged-in browser to automate daily tasks.

What the test found: Dependencies installed, project builds to dist/ (background.iife.js 1,848.81 kB, manifest.json), TypeScript type-check passes across all 15 packages, unit tests pass (14/14). 15 minutes.

61web-uiRuns90.3 / 100Python16,665

🖥️ Run AI Agent in your browser.

What the test found: The Browser Use WebUI serves HTTP 200 on port 7788 with the full Gradio UI, runs an agent against the local mock, and completes tasks with a decodable recording GIF and agent-history JSON. 37 minutes.

62OpenAdaptRuns90 / 100Python1,895

For EHRs, payer portals, and any app your team clicks through. A person or an agent shows the task; OpenAdapt turns it into a local program that repeats the...

What the test found: openadapt CLI installs and runs; tutorial flow records, compiles, certifies, and verifies a MockMed synthetic clinical workflow with VERIFIED receipts; --break-it correctly halts when the system of record rejects the write; MCP agent serve starts and serves workflows over stdio. 64 minutes.

63EndoplexityRuns90 / 100TypeScript774

Agentic browser control driven by your own Claude / Cursor subscriptions

What the test found: Bridge runs on Node 24, MCP endpoint responds with 14 browser tools, 175/175 tests pass, .endo-token file created with 0600 permissions. 3 minutes.

64nuphus-mcpRuns86.7 / 100Rust336

Desktop automation MCP server, computer use for any AI agent: control screen, windows, mouse/keyboard, and Chrome via Model Context Protocol (stdio). Not a DSH...

What the test found: nuphus-mcp v0.3.3 installs via npm, initializes over stdio JSON-RPC responding to initialize/tools/list/tools/call, returns screen size (1280x800), and captures valid PNG screenshots in headless Xvfb. 12 minutes.

65vmprintRuns80 / 100TypeScript681

This layout engine can solve the hardest layout problems imaginable. Its output is DTP grade and deterministic. It's faster than engines written in C++ because...

What the test found: vmprint CLI and engine ran in this container on Node 18 despite EBADENGINE install warnings: npm test exited 0 (engine 77+77 fixtures, packaged-integration 1 page/1107 bytes), the CLI rendered valid text-bearing PDFs with bundled local fonts and version 0.1.0, and engine.layout() produced 1 page x 1 box containing... 18 minutes.

66browser-useRuns66.7 / 100Python121,752

Open-source Python library for AI browser agents, plus Browser Use Cloud: stealth browsers and a hosted agent API.

What the test found: browser-use v0.13.10 installed in a Python 3.12 venv with uv, Chromium headless installed via playwright, and all 563 CI tests pass (529 passed, 34 skipped) after fixing 2 test identity comparison bugs in test_beta_agent.py that only manifest under xdist worker parallelism. 75 minutes.

67OpenCLIRuns66.7 / 100JavaScript31,225

Make Any Website into CLI & Use your logged-in browser by AI agent.

What the test found: OpenCLI 1.8.8 builds and runs on Node 22.14.0; 632 test files pass (7325 tests, 3 skipped) across unit, extension, adapter, smoke projects; CLI binary answers version 1.8.8 and lists 1366 registered commands. 27 minutes.

68stealth-browser-mcpRuns60 / 100Python3,385

Stealth browser automation for AI agents over MCP and the Chrome DevTools Protocol: navigation, network hooks, DOM extraction, and pixel-accurate UI cloning.

What the test found: Stealth Browser MCP server runs on stdio transport, spawns a headless Chromium 157.0.8086.0, navigates to data URLs, extracts page HTML content, and closes cleanly -- all via MCP protocol over stdio. 15 minutes.

69terminalRuns60 / 100Rust664

Terminal UI to get stuff done in the browser

What the test found: The Rust workspace builds cleanly, all 2058 Rust and 28 Python tests pass, cargo fmt is clean, and both the CLI (browser-use-terminal 0.1.8) and TUI (but 0.1.8) binaries run, with the TUI rendering its setup screen and responding to real keyboard input in a PTY session. 40 minutes.

70kitesurfRuns50 / 100TypeScript2,556

Kitesurf is an experimental browser engine that runs on Cloudflare Workers.

What the test found: The project builds from source, the dev server starts and responds to HTTP requests (health, root, /json/version), and at least one e2e test passes against the in-process engine. 34 minutes.

71Jobs_Applier_AI_Agent_AIHawkRuns with mocks92 / 100Python32,611

AIHawk: AI job application bot, automated job search, job hunting, auto-apply to resume-matching jobs while you sleep.

What the test found: The repository installs from source with all dependencies, the full automated suite passes (266 non-browser + 6 real-browser), and an end-to-end search-and-apply run through a mock OrcaRouter over a real local job index fills and submits a live Greenhouse application form, writing a decoded proof screenshot. 51 minutes.

72ego-liteRuns with mocks92 / 100JavaScript24,024

The fastest browser for AI agents to run browser automation, built for sharing your logged-in browser state with your AI agents, like Codex or Claude Code...

What the test found: 494 unit tests pass on Node 22.23.4 with mocking (FakeEgo test double), the build produces a CLI binary that prints help text, and type checking passes cleanly. Real browser e2e was not attempted because no ego lite runtime is available in this container. 2 minutes.

73GenericAgentRuns with mocks92 / 100Python14,441

Self-evolving agent: grows skill tree from 3.3K-line seed, achieving full system control with 6x less token consumption

What the test found: After a venv install of the declared UI and test dependencies, 275 tests pass, the TUI starts, the Streamlit UI completes a mocked tool-using turn with MOCK_FINAL: 246810, and launch.pyw opens a live 1200x900 GenericAgent window on DISPLAY=:99 using PyQt6 and an extracted libxcb-cursor. 29 minutes.

74llm-scraperRuns with mocks92 / 100TypeScript6,958

Turn any webpage into structured data using LLMs

What the test found: The llm-scraper repo installs (npm ci), builds with tsc, and its full vitest suite of 8 tests passes against a local schema-correct mock OpenAI server, with real Playwright-driven fetches of news.ycombinator.com and example.com. 49 minutes.

75google-qr-sms-helperRuns with mocks92 / 100JavaScript2,925

Chrome Extension (MV3) to switch Google's mandatory QR code verification back to standard 6-digit SMS verification via iPad Mini profile & [3,null,4] RPC...

What the test found: All 4 source files (manifest.json, background.js, content.js, ua_spoof.js) pass syntax validation. Chrome extension loads successfully via --load-extension with an active background service worker and a content script that injects a downgrade panel on accounts.google.com QR verification pages. 11 minutes.

76brightdata-mcpRuns with mocks92 / 100JavaScript2,733

A powerful Model Context Protocol (MCP) server that provides an all-in-one solution for public web access.

What the test found: All 21 unit tests pass with Node 20.20.2. The MCP server loads fastmcp without errors and rejects missing API_TOKEN. Dependencies installed cleanly. 2 minutes.

77lcuRuns with mocks92 / 100JavaScript2,456

Codex computer use, decoupled from the app, for usage inside any harness.

What the test found: LCU 0.9.1 builds into a thin release archive, installs into a prefix, passes all 435 Python unit tests, and answers version via the installed binary; 7 of 103 Node.js adapter tests fail on Node 18 due to MockTimers API version mismatch and missing .ts loader, both resolved by Node 20+. 11 minutes.

78actionbookRuns with mocks92 / 100Rust1,632

Let your AI agent get the sources behind logins and paywalls.

What the test found: Actionbook CLI v1.6.0 builds, reports version and help; MCP server starts and reaches the live API; JSON UI renders reports to HTML; SDK, MCP, tools-ai-sdk, openclaw-plugin, and json-ui packages all build and their 169 unit tests pass; the Rust CLI passes 482/483 lib tests (1 test requires lsof which is not available). 9 minutes.

79gemini-skillRuns with mocks92 / 100JavaScript835

gemini drawing MCP & skill through browser, can be used in openclaw or any agent that supports MCP. Gemini画图 MCP和sill,支持龙虾或任何agent使用٩(๑>◡<๑)۶

What the test found: Project installs, builds, and all infrastructure components function: daemon HTTP micro-service manages a headless Chrome instance via CDP, MCP stdio server exposes all Gemini/Atlas tools, atlas-provider connects to OpenAI-compatible APIs, and watermark-remover processes images through sharp - all tested end-to-end in... 4 minutes.

80surf-cliRuns with mocks92 / 100TypeScript706

The CLI for AI agents to control Chrome. Zero config, agent-agnostic, battle-tested.

What the test found: Surf v2.20.0 builds and all 81 test files pass (1177/1179 tests) using Node.js v22.12.0. The two test files that failed with SyntaxError were fixed by converting import-destructure patterns from require() to ESM import syntax. 19 minutes.

81agent-browser-cliRuns with mocks92 / 100Rust649

使用 agent-browser-cli 进行浏览器感知与控制。适用于标签页扫描/切换、页面 JS 执行、Cookie、CDP、contentSettings、截图、文件上传、下拉框点击、tmwd_cdp_bridge 初始化和 Web 工具排障

What the test found: Daemon runs on default ports with config matched; with the mock extension connected, status/doctor/tabs/tabtree/scan/exec/open/close/lookup/profile-label/network/console/screenshot/save-pdf/install-skill all return ok:true and doctor reports all checks ok, while click/fill/send-keys on @e refs and mouse-click remain... 35 minutes.

82jarvis-assistant-vocalRuns with mocks92 / 100Python571

Assistant vocal local en francais : Claude ou Ollama (offline), domotique Hue, OBS, agenda, navigateur, appels Twilio, serveur MCP. Python.

What the test found: Jarvis installs cleanly, its 181-test suite passes (4 Windows-only cases skipped), the Signal Matin pipeline emits valid A4 PDFs in all three density modes without overflow, and the OpenAI provider answers through a schema-correct local mock; only a physical microphone is missing. 36 minutes.

83browser-searchRuns with mocks92 / 100JavaScript550

A skill for AI agents: search the web with SearXNG, browse with Camofox, bypass protections with CloakBrowser. Anti-hallucination by design. Self-hosted, free...

What the test found: npm install succeeds, CloakBrowser (with Node 20, not the system Node 18) fetches live pages, the SearXNG and Camofox deterministic scripts pass their full command flows against schema-correct mocks, unit suites pass 28/28 and 12/12, and setup.sh/audit.sh complete without failed checks; real SearXNG/Camofox Docker... 33 minutes.

84flyto-coreRuns with mocks92 / 100Python483

AI said it finished. Flyto2 shows the proof.

What the test found: The Python package installs, builds, and 4751 unit/module/contract tests pass. The CLI responds to help. Documentation and brand checks pass. The container lacks system ping and a newer Node for jsdom@29, both worked around. 54 minutes.

85buxRuns with mocks92 / 100Python465

Browser Use Box: a 24/7 Claude Code agent for Playwright-style browser automation with Browser Use Cloud, Telegram, and a real browser on any box you own.

What the test found: Python venv at /tmp/bux-venv with websockets, httpx, Pillow installed. All 84 agent tests pass (telegram_bot: 63, mini_app: 16, bux_miniapp_tunnel: 3, login_flow: 2). The project can't fully launch without BROWSER_USE_API_KEY, TG_BOT_TOKEN, and systemd, those are real third-party services, but the full test suite... 4 minutes.

86open-browser-useRuns with mocks92 / 100JavaScript431

🔮 Platform-neutral Browser Use for AI agents: real Chrome automation with a CLI + SDKs, no lock-in, dead simple.

What the test found: The repo installs and builds in this container; the Go CLI/MCP/SDKs run and answer requests through a standing mock of the Chrome extension, and every test suite in the repo passes after the crypto fix. 13 minutes.

87herdr-browserRuns with mocks92 / 100TypeScript361

Render a real Chromium view inside a Herdr pane and drive it over CDP.

What the test found: Project builds, typechecks clean, all 140 tests pass, CLI help displays, and test HTTP server starts on localhost. 33 minutes.

88CatGPT-GatewayRuns with mocks92 / 100Python317

Turn your ChatGPT or Claude account into a fully working OpenAI-compatible API. No API keys needed. Supports tool calling, vision, file attachments, and image...

What the test found: CatGPT Gateway starts with PROVIDER=minimax, passes all 7 MiniMax provider tests, serves health/endpoints/models/chat completions on port 8000, and responds correctly to OpenAI SDK and curl requests. 6 minutes.

89resRuns with mocks92 / 100JavaScript314

Agent skill to reverse engineer a SaaS product through the browser into a knowledge base for rebuilding it: screens, rules, data model, design tokens and...

What the test found: All 32 automated tests pass under Node.js v22.14.0 and Playwright Chromium headless; all 4 CLI scripts exit 0 with correct usage and the end-to-end workflow (init, check, build-report, pack) produces valid output. 8 minutes.

90login-machineRuns with mocks92 / 100TypeScript292

AI-powered login automation. Uses Claude to classify login pages and Playwright to interact with them.

What the test found: The Next.js app builds, launches, returns HTTP 200 on / and /api/chat, and responds to all API actions (start, submit, close, error handling) using local Playwright Chromium with a mock heuristic page classifier. 17 minutes.

91opensteerRuns with mocks92 / 100TypeScript266

Opensteer turns your computer into an AI agent workstation that you can control from anywhere.

What the test found: Opensteer 0.11.2 is installed as a uv global tool, all 100 unit tests pass, the CLI responds to --version (0.11.2), --help, and -c execution, and skills are installed. 1 minute.

92pi-agent-browser-nativeRuns with mocks92 / 100TypeScript258

pi extension that exposes agent-browser as a native tool for browser automation

What the test found: npm install, build, and doctor pass; 1137 of 1256 tests pass with fake-agent-browser fixtures; 79 pre-existing cold-reopen and daemon-boundary failures remain. 84 minutes.

93agent-browser-mcpRuns with mocks92 / 100Python243

让 Agent 直接操作真实 Chrome 的 MCP 服务,支持页面扫描、CDP、截图与物理输入

What the test found: In a fresh .venv the project installs and builds, its CLI prints help/extension path/hermes config/doctor JSON, its MCP stdio server responds to initialize and tools/list with 19 tools, and against a mock Chrome bridge list_tabs, execute_js, scan_page, page screenshot, and desktop screenshot all returned real results... 28 minutes.

94dsh-antibrowRuns with mocks92 / 100JavaScript14

DeepSeek Harness plugin: give your agent a browser with a persistent identity - engine-level fingerprint spoofing, unlimited free local profiles, Android...

What the test found: The dsh-antibrow plugin package installs and packs cleanly; the smoke test passes with a mock MCP server implementing all 16 AntiBrow browser automation tools, verifying that the harness plugin bridge correctly discovers and routes all tools as mcp__antibrow__* entries. 11 minutes.

95HolyClaudeRuns with mocks86 / 100JavaScript2,680

AI coding workstation: Claude Code + web UI + 8 AI CLIs + headless browser + 50+ tools

What the test found: All 285 Node.js unit tests, 5 Python notification tests, and the CLI persistence shell test pass. Repository structure is validated end to end with product facts contracts, security policy, documentation, and patch scripts all verified. 2 minutes.

96cdp-bridge-mcpRuns with mocks86 / 100Python307

Let one AI workspace operate multiple signed-in Chrome accounts. Route each MCP request to its own browser profile through a self-hosted bridge.

What the test found: cdp-bridge installs cleanly, both stdio and streamable-http MCP transports start and respond correctly to initialize and tools/list, returning 12 browser automation tools ready for a Chrome extension connection. 5 minutes.

97agent-browser-mcpRuns with mocks86 / 100Python243

让 Agent 直接操作真实 Chrome 的 MCP 服务,支持页面扫描、CDP、截图与物理输入

What the test found: agent-browser-mcp CLI and MCP server install, start, and respond to all 20 tools; desktop automation tools work against the Xvfb display; browser CDP tools await a Chrome extension connection. 6 minutes.

98jev-socialRuns with mocks86 / 100JavaScript219

Open-source, local-first social media research agent for Instagram, TikTok, and LinkedIn. Jev routes read-only steps; socai CLI captures cited browser evidence.

What the test found: All 253 offline tests pass, the HTTP server starts and responds 200 on /, and the status JSON is correct. 3 minutes.

99ekoRuns with mocks80 / 100TypeScript4,962

Eko (Eko Keeps Operating) - Build Production-ready Agentic Workflow with Natural Language - eko.fellou.ai

What the test found: All 4 packages build successfully. 12 of 14 test suites pass (14 of 16 individual tests) in eko-core with a mock LLM server. Remaining failures: test/llm/utils.test.ts has a pre-existing unhandled rejection for an intentional timeout scenario; test/llm/mcp.test.ts requires an external MCP SSE server on localhost:8083... 27 minutes.

100dorabotRuns with mocks78.7 / 100TypeScript249

macOS app for 24/7 AI agents in an IDE with memory, scheduled tasks, browser use + access to Whatsapp, Telegram, Slack.

What the test found: The dorabot backend compiles from source, its WebSocket gateway serves HTTP health at 200 on a Unix socket and handles RPC methods for auth, sessions, skills, agents, calendar RRULE scheduling, goals CRUD, projects/tasks, and provider detection against a real SQLite (WAL) database; the interactive CLI lists skills and... 29 minutes.

101ClawBenchRuns with mocks76 / 100Python1,331

Open-source benchmark for browser AI agents on daily tasks.

What the test found: Python package clawbench-eval v0.10.0 installs and builds; all CLIs (clawbench, clawbench-run, clawbench-batch, clawbench-rescore, clawbench-analyze, clawbench-reproduce, clawbench-harbor-adapt, clawbench-edgebench-adapt, clawbench-edgebench-judge, clawbench-prorl-submit, clawbench-prorl-mock) are importable and... 3 minutes.

102nodRuns with mocks71 / 100TypeScript3,069

Open source personal agent with its own computer. It chases invoices, preps your calls and cancels trials on real websites, and asks before it acts. Runs on...

What the test found: Node 24.21.0 and pnpm 10.34.5 are installed; all dependencies resolved; TypeScript, architecture boundaries, and Prettier checks pass; 537/669 unit tests pass with 0 failures; the Vite web build produces the full dist bundle; the project is ready to run with Docker and model credentials. 33 minutes.

103parchiRuns with mocks56 / 100TypeScript548

Your AI friend right in your browser

What the test found: npm install completed 771 packages with warnings about node version requirements; npm run build produces dist/ with manifest.json, background.js, sidepanel/panel.{js,css,html}, and content scripts; npm test passes 233 checks; integration tests pass 19; orchestrator harness passes 7; tsc typechecks without errors... 12 minutes.

104Boterdrop-SolverRuns with mocks56 / 100Python229

Powerful CAPTCHA Solver (Turnstile, cf_clearance, reCAPTCHA v3, AWS WAF) built with FastAPI + Camoufox. High success rate, proxy support, and VPS friendly

What the test found: The FastAPI server runs on 0.0.0.0:8000, registers all 5 endpoints (/turnstile, /clearance, /aws-token, /recaptchaV3 GET+POST, /result), initializes a browser pool via camoufox, and correctly accepts task submissions and returns pollable results. 4 minutes.

105OpenBotCould not verify38.6 / 100TypeScript13,196

The open, self-hostable alternative to closed AI bots. Production-ready, fully customizable, and plugs into every stack: any agent, any model, any tool.

What the test found: bun install and all 3985 non-integration unit tests pass. The 69 failures are all infrastructure-gated (no PostgreSQL for integration tests, no Docker for compose tests, 2 pre-existing Mastra .env leakage failures). 49 minutes.

106browsenticCould not verify20 / 100TypeScript308

An agentic browser harness: an AI side panel & orb in your own logged-in Chrome, Edge, Arc, Brave or even Firefox run by the agent/subscription you already use...

What the test found: could not be verified; the log shows where it stopped. 8 minutes.

107pagesiftCould not verify20 / 100Rust303

Local-first web & document ingestion engine in Rust, HTTP/browser capture, layout-aware PDF/Office/OCR parsing, persistent sync, verifiable data graphs, and...

What the test found: could not be verified; the log shows where it stopped. 29 minutes.

-moliNot yet tested-Rust24,245-
-kane-cliNot yet tested-JavaScript267-
-playwrightNot yet tested-PHP261-
-safari-mcpNot yet tested-JavaScript248-

runs installed and started with its real dependencies. runs with mocks started after stand-ins replaced external services such as a database or a third-party API. could not verify neither the standard agent nor the stronger one got it running within the time limit; the log shows where it stopped.

How we tested

On this list as of the latest test: 70 projects ran as-is, 34 with mocks, 3 could not be verified, 4 still waiting. Languages tested: Go, JavaScript, PHP, Python, Rust, TypeScript. Every attempt used a clean single-use machine, the subject at a pinned version, and a 90-minute limit for standard projects; the complete procedure is on the methodology page.

Frequently asked questions (FAQs)

How is this list ranked?

By measurement, not opinion: projects Argusic installed and launched on a fresh machine come first, then those that ran with mocks in place of external services, then those it could not verify. Ties go to the Argusic Score, then how popular it is on its own source.

Why are some projects unranked?

4 projects are still waiting for a test or for a finished attempt. They are listed without a rank until Argusic has measured them.

Where is the evidence?

Every row links to the project's Argusic page, where each run has a full log and a terminal recording stored with a sha256 fingerprint. The same pages exist for every one of the tested projects, on this list or not.

More lists in this category

All lists: Best. All tested projects: subjects.