The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill uses a universal executor (run.js) that dynamically executes JavaScript code generated by the agent. The generated scripts are written to the /tmp directory and then loaded into the Node.js process using require(). This is a core feature for flexible browser automation but represents a high-capability execution primitive.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process content from external websites and local servers.
Ingestion points: Website navigation and data extraction functions in SKILL.md and lib/helpers.js (e.g., page.goto(TARGET_URL), extractTexts(page, selector)).
Boundary markers: None identified; the skill does not use specific delimiters or instructions to ignore embedded commands in the web content it reads.
Capability inventory: The skill environment allows for arbitrary JavaScript execution via run.js, which can access the file system, network, and spawn subprocesses via the agent.
Sanitization: No sanitization or filtering of the extracted web content is performed before it is returned to the agent's context.
[EXTERNAL_DOWNLOADS]: The skill's setup process and executor download the playwright library and browser binaries (Chromium) from official and well-known sources (npm and Playwright's official distribution channels).
[COMMAND_EXECUTION]: The skill executes shell commands to perform environment checks, detect running development servers on localhost ports, and manage browser installations via npm and npx.

playwright-skill