The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill includes run-code and eval commands that allow for the execution of arbitrary JavaScript within the browser context. This capability is documented for handling complex workflows, waiting strategies, and inspecting element attributes.
[DATA_EXPOSURE]: The skill provides tools for extracting sensitive browser state, including cookies (cookie-get), local storage (localstorage-get), and overall session state (state-save). The documentation notes that these should be handled carefully, but the feature allows for the direct retrieval of authentication tokens.
[EXTERNAL_DOWNLOADS]: The installation section suggests fetching code from the public NPM registry using npm install -g @playwright/cli@latest. This is a standard procedure for browser automation tools but involves remote dependency fetching.
[INDIRECT_PROMPT_INJECTION]: The skill navigates to and processes data from external, untrusted websites. 1. Ingestion points: The goto and open commands in SKILL.md and references/commands.md. 2. Boundary markers: Absent. 3. Capability inventory: The skill can execute browser code via run-code and read browser storage. 4. Sanitization: Relies on default browser sandboxing and Playwright's execution environment.
[DATA_EXFILTRATION]: The skill can capture network traffic, console logs, and visual recordings of the browser session via tracing-start and video-start commands, which may include sensitive data transmitted to or from the browser.

playwright-cli