The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill utilizes a custom executor (run.js) that dynamically writes JavaScript automation code to temporary files and executes them using the Node.js require() function. This mechanism is the core functionality used to perform custom browser tasks requested by the user.
[COMMAND_EXECUTION]: The skill executes shell commands for environment setup and infrastructure tasks. This includes running npm install and playwright install to manage dependencies, as well as executing server detection logic through the node run.js wrapper.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to navigate to and interact with arbitrary web pages provided by the user or discovered during development server detection. This creates a surface where malicious content on a target website could attempt to influence the agent's behavior during the automation process.
Ingestion points: User-provided URLs and web content retrieved during automation tasks in SKILL.md and lib/helpers.js.
Boundary markers: Absent; the instructions do not explicitly require the agent to use delimiters or ignore instructions embedded within the target web content.
Capability inventory: The skill possesses browser automation (Playwright), network request capabilities (http.request), and local file execution (require on generated scripts).
Sanitization: Absent; there are no explicit instructions or helper functions provided to sanitize or filter content extracted from external websites before it is processed by the agent.

playwright-skill