---

name: python description: Write and run Python scripts for data scraping, analysis, file processing, HTTP requests, and automation tasks. Python 3 is pre-installed with requests, beautifulsoup4, lxml, and playwright. allowed-tools: Bash(python*:*)

Python Environment

Python 3 is available in this container with pre-installed packages.

Pre-installed packages

• requests — HTTP client (GET, POST, APIs)
• beautifulsoup4 + lxml — HTML/XML parsing & web scraping
• playwright — Browser automation (shares system Chromium, no extra download)
• json, csv, sqlite3, re, datetime — standard library

Usage patterns

Quick one-liner

python3 -c "import requests; r = requests.get('https://example.com'); print(r.status_code)"

Write and run script

cat > /tmp/scrape.py << 'EOF'
import requests
from bs4 import BeautifulSoup

resp = requests.get('https://example.com')
soup = BeautifulSoup(resp.text, 'lxml')
print(soup.title.string)
EOF
python3 /tmp/scrape.py

Web scraping with BeautifulSoup

import requests
from bs4 import BeautifulSoup

url = 'https://www.set.or.th/th/market/product/stock/quote/SET/price'
headers = {'User-Agent': 'Mozilla/5.0'}
resp = requests.get(url, headers=headers)
soup = BeautifulSoup(resp.text, 'lxml')

# Extract data from tables
for row in soup.select('table tr'):
    cells = [td.text.strip() for td in row.select('td')]
    if cells:
        print(cells)

Playwright browser automation (for JavaScript-rendered pages)

from playwright.sync_api import sync_playwright
import os

with sync_playwright() as p:
    browser = p.chromium.launch(
        executable_path=os.environ.get('PLAYWRIGHT_CHROMIUM_EXECUTABLE_PATH', '/usr/bin/chromium'),
        headless=True,
        args=['--no-sandbox', '--disable-gpu']
    )
    page = browser.new_page()
    page.goto('https://example.com')
    page.wait_for_load_state('networkidle')
    content = page.content()
    browser.close()

HTTP API calls

import requests
import json

# GET with JSON response
data = requests.get('https://api.example.com/data').json()

# POST with payload
resp = requests.post('https://api.example.com/submit',
    json={'key': 'value'},
    headers={'Authorization': 'Bearer TOKEN'}
)

CSV/JSON file processing

import csv
import json

# Read CSV
with open('data.csv') as f:
    reader = csv.DictReader(f)
    for row in reader:
        print(row)

# Write JSON
with open('output.json', 'w') as f:
    json.dump(data, f, ensure_ascii=False, indent=2)

Installing additional packages at runtime

If you need a package not pre-installed:

pip3 install --break-system-packages pandas numpy
python3 script.py

Note: Runtime-installed packages are lost when this container exits. Pre-installed packages persist across runs.

Important notes

• Scripts should write output files to /workspace/group/ (persistent) not /tmp/ (lost on exit)
• For Playwright: always use headless=True and --no-sandbox args
• The system Chromium is shared between agent-browser and Playwright
• Prefer requests + BeautifulSoup for static pages (faster, lighter)
• Use Playwright only for JavaScript-heavy pages that need rendering

Browser Fallback Smoke Test

Use this when agent-browser is unavailable and you need to verify fallback quickly.

python3 - << 'PY'
from playwright.sync_api import sync_playwright
import os

with sync_playwright() as p:
    browser = p.chromium.launch(
        executable_path=os.environ.get('PLAYWRIGHT_CHROMIUM_EXECUTABLE_PATH', '/usr/bin/chromium'),
        headless=True,
        args=['--no-sandbox', '--disable-gpu'],
    )
    page = browser.new_page()
    page.goto('https://example.com', wait_until='domcontentloaded')
    print('title:', page.title())
    browser.close()
PY

If this succeeds, browser automation can continue via Python Playwright fallback.