Pilot Browser — AI skill for Claude Code
Let AI agents drive your real, logged-in browser: visible cursor, human takeover, approvals, injection-tested.
How to install Pilot Browser
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open ajstars1/pilot-browser and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Pilot Browser does
Let AI agents drive your real, logged-in browser: visible cursor, human takeover, approvals, injection-tested. MCP-native.
Alternatives in AI
- BrowserSkill — Let AI agents use your real, logged-in browser without interrupting your work 9.1k ★
- Browserwing — BrowserWing turns your browser actions into MCP commands Or Claude Skill, allowing AI agents to control browse 1.2k ★
- BossConsole — Open-source, multi-platform harness for AI agents — a native, multi-threaded operator's console (JVM, not Elec 227 ★
README
pilot-browser
**Let AI agents drive your real, logged-in browser, with a visible cursor, human takeover, and approvals for anything consequential. MCP-native.**
[](https://github.com/ajstars1/pilot-browser/actions/workflows/ci.yml) [](https://www.npmjs.com/package/@pilot-browser/mcp) [](LICENSE)

Real recording: Claude-style agent → pilot-browser MCP server → real Chrome, on a fictional job form (supervised mode). Re-record with `npm run demo:record`.
Most browser agents run a fresh headless browser with none of your logins, or ship a forked browser. pilot-browser attaches to the **Chrome, Brave or Edge you already use**, through the browser's own consent prompt (Chromium 144+). The agent works in a tab of its own; you watch its cursor move and can step in at any moment.
- Your browser, your sessions: no fork, no profile copying, no cookie export.
- You stay in control: clicking in the agent's tab pauses it. The agent can't see the page while you drive, so passwords and 2FA codes you type never reach the model.
- Approvals, at the level you choose: four modes from ask before everything to never ask. The default asks before submitting, paying, sending, deleting and uploading.
- Injection-resistant by construction: an origin allowlist, destination checks from the live DOM, cross-site copy checks and tamper-proof controls hold even when the model is fooled. They're tested against a fully compromised agent on every CI run (security model).
- MCP-native: works from Claude Code, Cursor, or any MCP client.
**Status: 0.1, early.** Tested on Linux with Chrome 154;
Related Skills
Chrome Bridge
CLI for browser automation: let any AI agent drive your real Chrome — logged-in tabs, zero deps, no relaunch,
Pilot
Chrome extension + MCP server — AI agents control a tab in your real browser, already logged in
Clawbrowse
Let Claude Code drive your real, logged-in Chrome. MV3 extension + zero-dep MCP server, element-table percepti
Pawbrowse
Let Claude Code drive your real, logged-in Chrome. MV3 extension + zero-dep MCP server, element-table percepti
Browser MCP
Drive your real, logged-in Chrome from any AI agent (Claude Code, Cursor, VS Code) — works where headless dies
Falaq Cockpit OSS
Mission control for Claude Code & Codex — a local dashboard to monitor, multitask, and drive all your AI codin
Related Agents
QA Desktop Tester
Desktop tester for flows outside the browser - native apps, OS dialogs, file pickers, Electron apps, OS settin
Aun Browser Runner
Executes a scoped browser packet in a fresh context, reads named fields, and returns evidence plus verified cl
LinkedIn Operator
Executes LinkedIn writes that have no API, through a logged-in browser MCP (Claude in Chrome or Playwright): p