Menü
Headless browser automation via agent-browser CLI
Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.
Basierend auf der SOC-Berufsklassifikation
Use when the user wants to create, inspect, update, or remove telclaude scheduled jobs, especially natural-language schedules like "every weekday at 9am". Prefer this skill for cron CRUD behind WRITE_LOCAL+.
Use when the operator asks for a daily or morning Telegram brief, agenda digest, inbox highlights, or Hermes-style operator preset. This is a read-only private-agent workflow that gathers today's Google Calendar, recent Gmail inbox highlights, and weather, then returns a concise Telegram brief.
Create, patch, pin, unpin, rename, or archive agent-authored telclaude skills through the guarded managed-skill writer. Use only for private/telegram agent work when durable skill changes are clearly useful.
Crafts Telegram-friendly replies for telclaude sessions, respecting media, heartbeats, and brevity.
Use when the operator asks for a weekly business report, Monday-morning business digest, revenue/support/CRM/analytics roundup, or wants to schedule workflow
{{description}}
| name | browser-automation |
| description | Headless browser automation via agent-browser CLI |
| allowed-tools | ["Bash","Read","Write","Glob","Grep"] |
Use agent-browser for headless Chromium browser automation when you need to:
# Navigate to a URL
agent-browser navigate --url "https://example.com"
# Take a snapshot (accessibility tree — lightweight alternative to screenshot)
agent-browser snapshot
# Take a screenshot (saves PNG — use cwd or /tmp)
agent-browser screenshot --path ./screenshot.png
# Click an element (by CSS selector or text)
agent-browser click --selector "button.submit"
agent-browser click --text "Sign In"
# Fill a form field
agent-browser fill --selector "input[name=email]" --value "user@example.com"
# Execute JavaScript in page context
agent-browser execute --script "document.title"
# Wait for an element to appear
agent-browser wait --selector ".results" --timeout 5000
All commands run in headless Chromium by default. No display server required.
Set PLAYWRIGHT_BROWSERS_PATH=/ms-playwright if browsers are installed there (Docker default).
agent-browser is not installed, fall back to WebFetch for simple page reads.--timeout to override.