The Agent Skills Directory

[PROMPT_INJECTION]: The skill instructions do not contain any attempt to bypass safety filters or override agent behavior. The YAML frontmatter and body are consistent with its stated purpose as a scraping tool.
[DATA_EXFILTRATION]: No sensitive file access or hardcoded credentials were detected. The network operations in scripts/playwright-simple.js and scripts/playwright-stealth.js are directed to user-specified URLs for the purpose of scraping content. Screenshots and HTML files are saved to local paths as requested by the user environment.
[EXTERNAL_DOWNLOADS]: The skill depends on the legitimate playwright package and uses standard installation procedures (npm install, npx playwright install chromium). No untrusted external scripts or binaries are executed.
[COMMAND_EXECUTION]: The skill executes shell commands to run the Node.js scraping scripts, which is the intended functionality. The scripts use Playwright's page.evaluate to interact with the DOM within the browser context, which is standard practice.
[SAFE]: The skill is a well-documented scraper. While it ingests untrusted data from the web (indirect prompt injection surface), it is designed for this purpose and outputs the data in a structured JSON format. No persistence mechanisms or obfuscation techniques were found.

smart-web-scraper