The Agent Skills Directory

[SAFE]: The skill is a legitimate technical resource for developers using Cloudflare's Browser Rendering service. It provides documentation, triggers, and templates for integrating headless Chromium automation into Cloudflare Workers.
[EXTERNAL_DOWNLOADS]: The skill references official documentation and package repositories from well-known and trusted sources, including Cloudflare's own developer documentation, the official Puppeteer (Google) and Playwright (Microsoft) project sites, and associated GitHub repositories.
[INDIRECT_PROMPT_INJECTION]: The skill deals with processing external, untrusted web content and URLs through browser automation, which is its primary functional purpose. While this involves an inherent risk of indirect prompt injection, the skill includes a detailed 'Production Checklist' that explicitly recommends security mitigations such as input validation for URLs to prevent SSRF, timeout limits, and authentication for scraping endpoints.
Ingestion points: page.goto(url) and page.setContent(html) are used in templates within SKILL.md to process external data.
Boundary markers: While not present in the code snippets themselves, the documentation instructs users to implement authentication and input validation.
Capability inventory: The skill utilizes page.evaluate() for browser-side JavaScript execution, page.screenshot(), page.pdf(), and general network navigation capabilities.
Sanitization: The documentation recommends input validation for URLs as a production requirement.

cloudflare-browser-rendering