The Agent Skills Directory

[SAFE]: The skill provides legitimate instructions for UI automation and verification using well-known browser testing methodologies. It focuses on ensuring that coordinate-based interactions result in expected application state changes.- [COMMAND_EXECUTION]: Includes example commands for playwright-cli and kill-port. These are standard tools for web development and automated testing environments and are used here within their intended scope for local development verification.- [INDIRECT_PROMPT_INJECTION]: The protocol involves reading browser snapshots and console errors from the application under test. While this exposes the agent to data from external web contexts, the instructions are directed toward testing the user's own project during development, which is a standard and expected behavior for this use case.
Ingestion points: Browser snapshots and console output (SKILL.md)
Boundary markers: Not explicitly defined for tool output processing
Capability inventory: Command execution via playwright-cli (SKILL.md)
Sanitization: Not explicitly defined for processed browser data

dnd