The Agent Skills Directory

[COMMAND_EXECUTION]: The skill makes extensive use of the Bash tool to execute playwright-cli, npm, and npx commands for browser automation and package management.
[EXTERNAL_DOWNLOADS]: The instructions include steps to install the Playwright CLI and initialize Playwright projects using standard package managers (NPM/NPX), which are well-known and trusted developer tools.
[DATA_EXPOSURE]: The skill provides commands to list and retrieve sensitive browser data, including cookies, localStorage, and sessionStorage. This is a primary feature for session management and automation testing.
[DYNAMIC_EXECUTION]: The run-code and eval commands allow for the execution of arbitrary JavaScript within the browser context. This capability is used to handle advanced scenarios like geolocation overrides, custom wait strategies, and complex web interactions.
[INDIRECT_PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it ingests untrusted data from web pages (DOM snapshots, titles, attributes) which are then processed by the agent.
Ingestion points: playwright-cli snapshot, playwright-cli eval, and playwright-cli show --annotate (SKILL.md).
Boundary markers: None identified in the prompt interpolation process.
Capability inventory: Subprocess calls via Bash tool and arbitrary JS execution via run-code (references/running-code.md).
Sanitization: No explicit sanitization or filtering of external web content is described.

playwright-cli