Browser Cash

**Judgement:** High-capability, advanced browser automation skill that leverages Browser.cash to bypass anti-bot protections and expose a CDP WebSocket endpoint for Playwright/Puppeteer. Technically strong and flexible, but carries significant abuse and policy risk if misused. **What it does well** - Spins up remote, “real” browser sessions that are harder to detect than typical headless automation. - Offers geo-targeting, window sizing, and persistent profiles for realistic, stateful sessions. - Provides clear, end-to-end examples for session creation, interaction via Playwright/Puppeteer, status checking, listing, and cleanup. - Integrates with Clawdbot’s config system for API key management and shows how to ensure Node/Playwright dependencies are present. **Key risks / concerns** - Explicitly designed to bypass anti-bot protections; misuse could violate site ToS, data protection laws, or anti-abuse policies. - Enables large-scale scraping and automation on protected sites, which could be repurposed for credential stuffing, aggressive scraping, or other hostile automation if a user is malicious. - Relies on an external paid service (Browser.cash); API key leakage or misconfiguration could lead to unintended usage and cost. - Operational reliability depends entirely on Browser.cash availability and network stability. **Recommended scenarios** - You control or have explicit permission to test/scrape a site that is behind Cloudflare/DataDome/etc., and need realistic browser behavior for QA, monitoring, or feature tests. - You need robust scraping or automation of complex, JS-heavy pages where standard headless browsers are blocked but ToS-compliant automation is allowed. - You require long-lived, stateful browser sessions (e.g., staying logged into your own accounts) with persistent profiles. **Not recommended / use with extreme caution** - Automating access to sites that forbid scraping or automated access in their ToS. - Any scenario that could be construed as evading security controls without explicit authorization. - High-volume scraping without rate limiting or respect for robots.txt and legal constraints. Overall, this skill is powerful and well-documented for experienced users who understand CDP tooling, but it should only be used in strictly legal, permissioned contexts with careful handling of API keys and resource usage.