OpenClaw Web Automation (No Credentials)

This skill is the fastest path for public-site automation tasks.

Use this for requests like:

• "Summarize https://www.yahoo.com"
• "Check whether https://example.com mentions pricing"
• "Count mentions of 'news' on a public page"

Do not use this skill for login-required tasks. Use award/login flows instead.

Preconditions

• Repository is available locally.
• Python environment is installed (pip install -e .).
• No credentials are required for this skill.

Command to run

From repo root:

python skills/openclaw-web-automation-basic/scripts/run_query.py --query "<user request>"

The script routes to examples/public_page_check and returns JSON.

Output handling

• Return a concise summary to the user:
  • page title
  • keyword
  • keyword count
  • 1-3 highlights
• If the query has no URL, default to https://www.yahoo.com.
• If parsing fails, ask the user for a URL and what text to check.

Safety

• Public websites only.
• No credential collection.
• No anti-bot bypass guidance.