The Agent Skills Directory

[DYNAMIC_EXECUTION]: The run.js script implements a 'Universal Executor' pattern. It takes arbitrary code (either from an argument, a file, or stdin), wraps it in an asynchronous template, writes it to a temporary file (.temp-execution-*.js), and then executes it using require(). This allows for the execution of arbitrary JavaScript within the Node.js environment.
[COMMAND_EXECUTION]: The run.js script uses execSync to perform environment setup, specifically running npm install and npx playwright install chromium. These commands are executed automatically if Playwright is not detected.
[EXTERNAL_DOWNLOADS]: The skill downloads and installs external dependencies and browser binaries.
Fetches the playwright package from the official NPM registry.
Downloads the Chromium browser binary via npx playwright install during the setup phase.
[INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it processes user-provided descriptions of web tasks to generate executable code.
Ingestion points: User descriptions of automation tasks (e.g., 'Test this login page') provided via the agent prompt.
Boundary markers: None identified in the provided files to separate user instructions from system code generation.
Capability inventory: The skill has the capability to perform network requests (via Playwright), write files (via run.js and Playwright screenshots), and execute arbitrary JS code (via run.js).
Sanitization: There is no evidence of sanitization or validation of the generated code before it is written to disk and executed.
[METADATA_POISONING]: There is a discrepancy between the author 'foraefactory' specified in the analysis context and the author 'lackeyjb' listed in the package.json file. This mismatch can be deceptive regarding the skill's origin.

playwright-skill