The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill generates and executes a Node.js script (analyze.js) to automate browser interactions and analyze web pages.
Evidence: The script content and execution instructions are defined in references/playwright-analysis.md.
[EXTERNAL_DOWNLOADS]: The skill downloads and installs the playwright package and Chromium browser binary via npm and npx.
Evidence: Commands in references/playwright-analysis.md install these dependencies from official registries.
[COMMAND_EXECUTION]: The skill uses shell commands to prepare the environment and run the analysis script.
Evidence: The instructions in references/playwright-analysis.md include shell commands like mkdir, npm install, and node.
[PROMPT_INJECTION]: The skill has an indirect prompt injection surface because it ingests and processes text content from external URLs provided by the user.
Ingestion points: The script in references/playwright-analysis.md captures document.body.innerText from external websites.
Boundary markers: Absent. There are no delimiters or specific instructions to ignore embedded commands within the ingested content.
Capability inventory: The agent uses the scraped text and screenshots to perform conversion rate optimization (CRO) audits and provide recommendations.
Sanitization: Absent. The visible text and DOM data are processed without filtering or validation.

popups