Ganidhuz-FoxX 🦊

Browse X/Twitter through Firefox with your real session cookies. Built because Chromium kept getting bot-blocked by X.

Requirements

• Python 3.7+
• Playwright: pip install playwright && playwright install firefox
• Firefox installed with an active X/Twitter login
• Xvfb display (for headless servers): Xvfb :1 &

Setup

1. Export your X cookies

Close Firefox first, then:

bash scripts/export-x-cookies.sh
# Cookies saved to secrets/x-cookies.json by default
# Override: FOXX_COOKIES_OUT=/custom/path.json bash scripts/export-x-cookies.sh

Custom Firefox profile path:

FIREFOX_PROFILE_PATH=/path/to/profile bash scripts/export-x-cookies.sh

2. Health check

bash scripts/check-firefox-env.sh

Usage

Run a plan file:

DISPLAY=:1 python3 scripts/playwright-firefox-control.py --plan /tmp/foxx-plan.json

Plan Examples

View a profile

{
  "needs_gui": true,
  "gui_reason": "site_only_action",
  "url": "https://x.com/elonmusk",
  "cookies_path": "secrets/x-cookies.json",
  "steps": [
    {"action": "wait", "ms": 4000},
    {"action": "screenshot", "path": "/tmp/foxx-profile.png"}
  ],
  "close_delay_ms": 3000
}

Search tweets (live)

{
  "needs_gui": true,
  "gui_reason": "site_only_action",
  "url": "https://x.com/search?q=AI+agents&src=typed_query&f=live",
  "cookies_path": "secrets/x-cookies.json",
  "steps": [
    {"action": "wait", "ms": 4000},
    {"action": "screenshot", "path": "/tmp/foxx-search.png"}
  ],
  "close_delay_ms": 3000
}

Fetch a tweet

{
  "needs_gui": true,
  "gui_reason": "site_only_action",
  "url": "https://x.com/user/status/123456789",
  "cookies_path": "secrets/x-cookies.json",
  "steps": [
    {"action": "wait", "ms": 3000},
    {"action": "content", "selector": "article"},
    {"action": "screenshot", "path": "/tmp/foxx-tweet.png"}
  ],
  "close_delay_ms": 3000
}

Plan Options

Supported Step Actions

• goto - navigate to URL
• click - click element by selector
• fill - fill input by selector
• type - type text with delay
• press - press keyboard key
• wait - wait ms
• wait_for_selector - wait for element
• screenshot - take screenshot
• content - extract inner text from element

Behaviour Rules

• Always wait for page load (wait step, min 3000ms recommended)
• A validation screenshot is always taken before closing
• Browser waits close_delay_ms before closing - verify result is correct
• If cookies expired (redirects to login), re-run export-x-cookies.sh
• Always close browser after task - don't leave it idle