Loading repository data…
alexvilelabah / repository
Open-source AI browser agent — type plain-language commands and Bah operates the web for you. Works with free AI, cloud providers, or local Ollama models.
A transparent discovery signal based on current public GitHub metadata.
This score does not audit code, security, maintainers, documentation quality, or suitability. Verify the repository and its current documentation before adoption.
You give natural-language commands ("open gmail and delete the spam") and the AI operates the browser in your place — reading the screen, clicking with a real mouse, typing, and going until it's done.
💸 Free out of the box — no key, no GPU. Chat and image generation work the moment you open it, with no API key. For the full autonomous agent (or more speed), add a cheap key — DeepSeek (recommended), Mistral or NVIDIA NIM. Prefer 100% local + offline? Run a model with Ollama.
▶️ Download the full video — the agent searching and operating the web on its own. (the GIF above is a preview; GitHub doesn't play embedded video in the README.)
🧑 I just want to use it (Windows): grab the installer here → the Bah-Setup-*.exe file, double-click and install.
🏠 Want it 100% local & free? No terminal needed. Set up the AI right inside the browser: open the AI panel → 🏠 Local AI → type a model name → Download, and Bah pulls the Ollama model for you. (Prefer the cloud? A cheap DeepSeek key works too — and needs no GPU.)
🔄 Auto-updates: after installing, Bah checks for new versions, downloads them in the background and offers "Restart now" to apply — no reinstalling.
⚠️ Windows shows a blue "protected your PC" screen (the app isn't code-signed with a paid certificate yet). Click More info → Run anyway — normal for new open-source apps. (Later updates install without that warning.)
👨💻 I want to hack on the code: clone and run it — see Running it below.
Selected from shared topics, language and repository description—not editorial ratings.
suitedaces /
macOS app for 24/7 AI agents in an IDE with memory, scheduled tasks, browser use + access to Whatsapp, Telegram, Slack.
83/100 healthmits-pl /
Open-source AI coding agent. Desktop app, bring your own model. Writes code, browses the web, verifies its work. Apache 2.0.
58/100 healthuiuing /
sendInputEvent (not synthetic events — goes through React, Vue, Angular without being ignored)navigator.webdriver) so sites don't wrongly block the browser| Layer | Tech |
|---|---|
| Browser shell | Electron 42 + Chromium |
| UI | React 19 + TypeScript + Vite |
| AI (cloud) | DeepSeek (recommended) · Mistral · NVIDIA NIM · free no-key fallback (Pollinations) |
| AI (local) | Ollama |
| Adblock | @ghostery/adblocker-electron |
| Webview | <webview> tag with persistent partition |
USER → "open gmail and delete the spam"
│
▼
┌───────────────────────────────────────────────┐
│ for step in 1..25: │
│ 1. observePage(webview) │
│ → { url, title, interactive_elements } │
│ 2. captureScreenshot() │
│ 3. AI decides ONE action: │
│ { action: { type, ...params } } │
│ 4. execute action via REAL OS input │
│ 5. wait, re-observe, self-evaluate │
│ 6. if action == 'done' → return │
└───────────────────────────────────────────────┘
| Action | What it does |
|---|---|
click_ref(N) | Clicks the element with id N from the observed list |
fill_ref(N, value) | Fills input id N with value (and verifies it took) |
click_text(text) | Finds by visible text and clicks |
click_at(x, y) | Click at exact coordinates (visual fallback) |
type(text) / press(key) | Type into the focused field / send a key |
navigate(url) / scroll(dir) | Go to a URL / scroll |
new_tab / switch_tab / close_tab | Tab management |
done(reason, success) | End the loop |
Clicks happen through webContents.sendInputEvent in the main process — a real Chromium mouse event, not a synthetic one, so React/Vue/anti-bot sites respond normally. The AI prefers the DOM-first path (click_ref), falling back to text and then coordinates.
git clone https://github.com/alexvilelabah/bah-browser.git
cd bah-browser
npm install
npm run build
npm start # or: npx electron .
Windows shortcut: double-click Abrir-Bah.bat.
127.0.0.1:11434). Then download a model from inside Bah (☁️/🏠 → 🏠 Local AI → type a name → Download) or in a terminal (e.g. ollama pull qwen3:14b). Local works offline, but the cloud (DeepSeek) is more reliable.The agent runs with full browser privileges, so it's worth being clear about what it does and doesn't do:
⚖️ You're in control — and responsible. Bah acts in your real session, on your account. Use it within each site's terms and the law. Sensitive actions (paying, buying, deleting, entering card data) always ask for your confirmation first.
🔍 Audit it yourself. The full threat model, a verify-it-yourself checklist (each protection → the exact file), what leaves your machine, and an honest list of accepted tradeoffs all live in SECURITY.md. A map of the codebase — every file, the live agent loop, the build — is in ARCHITECTURE.md. The whole of
src/is ~30 files.
🔓 It's your real session. The browser uses a persistent partition (persist:browser), so cookies and logins are saved. If you're logged into Gmail in Bah, so is the agent. The AI can access anything you could access manually. Don't log into accounts you wouldn't trust an assistant with.
🛡️ Safety brake on sensitive actions. Before paying, buying, deleting, or entering card data, the agent pauses and asks for your confirmation — and this works on every path (model clicks, coordinate clicks, Enter on a checkout page, learned shortcuts, and repeated automations). It never does those silently.
🛑 Stop means stop. The ■ Stop button cancels immediately, even mid model-call or mid-loop; a late response won't "resurrect" a cancelled task.
🎯 No fake success. After a fill/type the agent checks the field actually holds the value; if an action had no real effect it switches strategy instead of reporting success.
🙋 Asks for help when blocked. On a CAPTCHA, login wall, or paywall it stops and asks you to step in, then resumes — it doesn't flail.
🔢 25-step cap per command. If a task doesn't finish in 25 actions, the agent stops on its own.
🧩 Compatibility, not evasion. We present as standard Chrome (Chrome UA + masking navigator.webdriver) only to avoid being wrongly blocked. We do not break CAPTCHAs, dodge rate-limits, or automate things sites forbid in their terms.
🔑 Google login — use the "Sign in to Google" button. Google blocks login inside embedded browsers (Electron/webview). Bah handles it the right way: click 🔑 Sign in to Google → it opens the login in your real Chrome/Edge (where Google trusts it), you sign in, and Bah detects it automatically, imports the session (cookies via CDP) and closes the login window. Do it once and you stay logged in.
🚫 Adblock pauses on known sites. YouTube and Twitch get automatic bypass so their player isn't blocked by anti-adblock. Everywhere else adblock stays on.
Not implemented yet (but would be nice): a "preview the plan, then approve" mode before running, a per-tab sandbox, and click rate-limiting to avoid aggressive bot-like behavior.
| Bah | Comet | Browser-Use | |
|---|---|---|---|
| Open source | ✅ | ❌ | ✅ |
| 100% local option (Ollama) | ✅ | ❌ | ✅ |
| Runs at home | ✅ | ❌ | ❌ (lib only) |
| Cloud or local AI | ✅ | ❌ (cloud only) | ✅ |
| Real clicks (not synthetic) | ✅ | ✅ | ✅ |
| Full UI | ✅ | ✅ | ❌ |
| Adblock + Safe Browsing | ✅ | ✅ | ❌ |
| Confirmation before sensitive actions | ✅ | ⚠️ | ❌ |
ℹ️ The tested and recommended AI path is DeepSeek (cloud); local (Ollama) also works but is less validated.
MIT — see LICENSE.
In short:
An open-source AI agent for the browser. Bring any model. It plans in the side panel, acts on real pages through tools, and verifies every claimed result against the live DOM — with guardrails and reusable skills.
manushi4 /
Give AI eyes and hands on your desktop. Open-source MCP server for desktop automation — screenshots, UI control, browser automation, OCR. Works with Claude, Cursor, and any MCP client. macOS + Windows.
57/100 healthtawgroup /
Give any AI a real computer. Open source MCP server with Ubuntu sandbox, browser automation, desktop control, and 30+ tools. Works with Claude Code, Cursor, Claude Desktop.
67/100 healthLiamFuller07 /
Open-source alternative to Claude Cowork - Universal AI agent platform for completing tasks with AI. Model-agnostic (Claude, Gemini, OpenAI, Ollama), plan-mode by default, sandboxed file ops, browser automation.
47/100 health