The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands like 'lsof', 'curl', and 'playwright-cli' to interact with the local environment and web services. This behavior is consistent with the stated purpose of verifying web UI and server status.
[EXTERNAL_DOWNLOADS]: The skill uses 'npx -y playwright' to download and execute Playwright tools from the NPM registry, a well-known service. This is a standard method for ensuring the latest version of a tool is available during execution.
[INDIRECT_PROMPT_INJECTION]: The skill processes external data by capturing accessibility trees and snapshots from web pages. While this creates a potential attack surface for indirect prompt injection, it is a functional requirement for the skill's purpose. Ingestion points: Playwright snapshot and screenshot data (SKILL.md). Boundary markers: None explicitly mentioned in the instructions to distinguish page content from instructions. Capability inventory: Shell execution via 'lsof', 'curl', 'npx', and 'playwright-cli' (SKILL.md). Sanitization: No specific sanitization or filtering of external web content is described.

browser-verify